The mneme

Cornell University
Library

The original of tiiis book is in
tine Cornell University Library.

There are no known copyright restrictions in
the United States on the use of the text.

http://www.archive.org/details/cu31924100387210

3 1924 100 387 210

In compliance with current

Copyright law, Cornell University

Library produced this

replacement volume on paper

that meets the ANSI Standard

Z39.48-1992 to replace the

irreparably deteriorated original.

2005

CORNELL

UNIVERSITY

LIBRARY

GIFT OF
Professor H. S. Liddell

BF703 .S4°7"l92l""'""' '"""'''
Mneme.

olin

• ^ 10^4 olii y^s 4e e-

THE MNEME

THE MNEME

BY

RICHARD SEMON

LONDON: GEORGE ALLEN & UNWIN LTD.
RUSKIN HOUSE, 40 MUSEUM STREET, W.C. i

NEW YORK: THE MACMILLAN COMPANY

f *

•X

^.. ..<^

^

F»w/ published in igii

(All rights reserved)

PREFACE

In translating the Mneme during the years 1912-14 I
had the benefit of the constant advice of its distinguished
author, whose great wish it was to present his theory,
in English, to the scientific world.

The translation was completed in July 1914, but the
publication of the work could not be arranged for until
six years later, after the death of the author.

Professor Semon died at Munich in 1919, in his sixty-
first year.

A sequel to the Mneme {Die mnemischen Empfindungen)
was published in 1909 by Wilhelm Engelmann, Leipzig.
A posthumous work, Bewusstseinsvorgang und Gehirn-
prozess, has just been published by Bergmann,
Wiesbaden.

I desire to acknowledge my indebtedness to the Rev.
Lawrence Schroeder, M.A., of Halifax, and Mr. L. H.
Edminson, M.A., of London, for their valuable help in
revision.

LOUIS SIMON
Altrincham, Cheshire,
January, 1921.

THE ENGLISH EDITION OF

THE MNEME

IS DEDICATED TO

SIR FRANCIS DARWIN

CONTENTS

INTRODUCTION

PART I

CHAPTBK

I. ON STIMULUS AND EXCITATION . . 17

II. ENGRAPHIC ACTION OF STIMULI ON THE

INDIVIDUAL . . . , .24

in. ENGRAPHIC ACTION OF STIMULI ON PROGENY $7

PART II

IV. THE MUTUAL RELATIONS OF ENGRAMS: SIMUL-

TANEOUS AND SUCCESSIVE ASSOCIATION . 89

V. THE LOCALISATION OF ENGRAMS . . II3
VL ECPHORY OF THE EN GRAM. THE TWO

PRINCIPAL MNEMIC LAWS . . . 138

VII. MNEMIC EXCITATION AND HOMOPHONY . I49

PART III

Vin. EVIDENCE OF THE MNEMIC FACTOR IN ONTO-
GENETIC REPRODUCTION . . . 173

IX. THE INITIAL ONTOGENETIC ENGRAM AND THE

ONTOGENETIC COURSE OF DEVELOPMENT . 191

8 THE MNEME

CHAPTER PAGE

X. MORPHOGENEOUS MNEMIC EXCITATIONS IN

THE FULLY DEVELOPED ORGANISM . . 202

XI. ENGRAM LOCALISATION AND REGENERATION . 207

XII. ENGRAM DICHOTOMY IN ONTOGENESIS. . 221

XIIL THE ORIGIN OF ONTOGENETIC ENGRAM-

SUCCESSIONS . . . . . 240

XIV. ENGRAPHIC ORIGIN OF THE DETERMINANTS . 256

XV. THE PROPORTIONAL VARIABLENESS OF MNEMIC

EXCITATIONS ..... 266

PART IV

XVL RETROSPECT— REJOINDER TO CRITICISMS . 273

XVII. THE MNEME AS THE PERSISTING ELEMENT IN

THE CHANGE OF ORGANIC PHENOMENA . 288

INDEX OF LITERATURE .... 295
INDEX OF AUTHORS . . . 30O

GENERAL INDEX ..... 302

INTRODUCTION

The attempt to discover analogies between the various
organic phenomena of reproduction is by no means new.
It would have been strange if philosophers and naturalists
had not been struck by the similarity existing between
the reproduction in offspring of the shape and other
characteristics of parent organisms, and that other kind
of reproduction which we call memory. Should the
present or any subsequent author succeed in proving that
this similarity is more than superficial, there will be
no lack of critics who will remind us that some ancient
or modern thinker has already conceived the idea. Have
we not already heard heredity described as a kind of
racial memory (memory inherent in the species) ?

The first serious enquiry into the nature of this coin-
cidence was made by the celebrated physiologist Ewald
Hering in a paper read on May 30, 1870, before the
Vienna Academy, entitled, " Memory, a Universal Function
of Organic Matter." In this short paper of only twenty
pages Hering, with admirable insight and clearness,
summed up the chief points of resemblance between the
reproductive powers of heredity, of practice and habit,
and of conscious memory.

Satisfied, however, with combining these in a har-
monious scheme — thus pointing out the way for future
enquiry — Hering refrained from analytically demon-
strating that the resemblances between the different
reproductive processes were more than accidental, and
left behind him the task of proving that all these repro-,
ductive processes — whether of heredity, habit, or memory
— owe their resemblance to their common origin in onel
and the same faculty of organic matter.

10 THE MNEME

A few years later the English physician, F. Laycock,
evidently unaware of Hering's paper, elaborated a similar
thesis in an interesting essay entitled, " A Chapter on
Some Organic Laws of Personal and Ancestral Memory "
(Journal of Mental Science, vol. xxi, 1879).

A more explicit treatment of the problem was attempted
in 1878, in Life and Habit, by Samuel Butler, the well-
known author of Erewhon. Butler endeavoured to trace
analogies between the various reproductive processes
in greater detail than had been attempted by Hering,
whose paper came to Butler's knowledge only after the
appearance in 1880 of Unconscious Memory, his own
first publication on the subject. Butler's essay contains
brilliant suggestions, but these are mixed with so much
questionable matter, that the whole, compared with
Hering's paper on the same subject, is rather a retrogression
than an advance.

Many years later, similar ideas were set forth by Henry
D. Orr in a book entitled, A Theory of Development and
Heredity {New York, 1893), and with as complete independ-
ence of Hering as was the case with Laycock and Butler.

Of these four independent attempts to connect the
phenomena of memory, heredity, and habit, not one
seems to have had any real influence on contemporary
scientific thought, for they all omitted to explain definitely
why such dissimilar processes should possess in common
an obviously repetitive nature.

Mere repetition is of course not confined to these three
classes of phenomena, nor yet even to the organic king-
dom. Repetition of similar modes and peculiarities occurs
frequently in the inorganic world without our ever being
tempted to compare such repetition with the phenomena
of memory. Indeed, we expect repetition of this kind,
that is to say, of modes and qualities, whenever there
is a repetition of the original conditions. But the
pecuUarity of memory is that its kind of repetition
is independent of a complete repetition of the original
conditions.

The periodic eruptions of a geyser, or the recurrence

INTRODUCTION ii

of spring tides each time a given lunar phase and a given
position of the sun coincide, do not suggest any com-
parison with the phenomena of memory. We simply
recognise that when all the specific conditions are present,
then, and only then, certain phenomena are invariably
repeated. But if we wish to distinguish from these the
various phenomena of organic reproduction, such as
memory, habit, periodicity, and heredity, as a group
by themselves, we must show that these latter possess a
common quality which differentiates them from all other
kinds of repetition. This characteristic they possess in
the capacity for repetition, even when the original con-
ditions are varied to an extreme degree. An infinitesimal
portion, indeed, of the original conditions often suffices to
cause the reproduction.

To show this demands a thorough analysis of these
phenomena, not only in all their details and connections,
but also in their verbal and logical formulation. As such
an analysis is necessarily a complicated piece of work,
requiring considerable patience and attention on the
part of the reader, it may be well to acquaint him with
the aim and scope of the following chapters, and to
present a general outline of the problems discussed. He
must also be famiharised with some of the words I have
coined in the hope of avoiding the misleading connotations
inherent in everyday speech.

First of all I wish to point out that, instead of speaking
of a factor of memory, a factor of hahit, or a factor of
heredity, and attempting to identify one with another,
I have preferred to consider these as manifestations of
a common principle, which I shall call the mnemic
principle. This mnemic property may be regarded from
a purely physiological point of view, inasmuch as it is
traced back to the effect of stimuli applied to the irritable
organic substance. But the immediate effect of stimulation
on the irritable substance is only one half of the problem
with which we are concerned, although it happens to
be that which has mainly occupied the attention of
investigators. The other and distinctive half of the

12 THE MNEM£

mnemic problem underlying the problems of memory,
habit, and heredity, is the effect which remains in the
stimulated substance after the excitement produced by
the stimulation has apparently ceased. The capacity
for such after-effect of stimulation constitutes what I
have called the Mneme. Its result, namely, the enduring
though primarily latent modification in the irritable sub-
stance produced by a stimulus, I have called an Engram,
and the effect of certain stimulations upon certain
substances is referred to as their Engraphic effect.

Now, it is attested by numerous observations and
experiments that the engraphic effects of stimulation are
not restricted to the irritable substance of the individual,
but that the offspring of that individual may manifest
corresponding engraphic modifications. Nor is there
anything surprising in this, once we recognise not only
that the offspring is produced from the germ-cells of
the generation which has been submitted to the particular
engraphic stimulation, but that these germ-cells are in
continuous organic connection with the rest of the irritable
substance of the organism. This once demonstrated,
it should not be difficult to prove that the engraphic
effects of stimulation are subject to the same laws,
whether these effects are manifested in the individual
which originally experienced stimulation, or in what we
call the particular inherited character of that individual's
offspring.

So much for the general nature of engrams, engraphic
action, and effects, and of the principle which has given
the name of Mneme to this book.

The second part of the book endeavours to trace all
mnemic phenomena to a common origin. In doing this
we deal with the influences which awaken the mnemic
trace or engram out of its latent state into one of mani-
fested activity, a process to which is given the name of
Ecphory. Our study reveals the laws regulating the
various associations between the latent and the revived,
or ecphorised, mnemic traces or engrams. Going still
further into the subject, we deal with the manner in

INTRODUCTION 13

which the stock of engrams has been originally ac-
quired, and also with the way in which this stock
has been partially transmitted by heredity. Mean-
while, the reader is given some notion of the con-
cordant action of closely allied mnemic and original
excitations, a consonance which I have found it convenient
to call Homophony. Broadly, the aim at this point in
the enquiry is to reduce the phenomena of individually
acquired reminiscences and the phenomena of individually
acquired habit to the same common mnemic elements
as are discernible in the large group of reproductive
phenomena usually referred to heredity. These common
mnemic elements are dealt with in special reference to
their capacity for making and reviving impressions
(Engraphy and Ecphory). Everyday expressions which
have hitherto led to many mistaken verbal analogies are
thus excluded.

The third part of the book seeks to furnish concrete
proof that the development of the individual ontogeny
is explicable by mnemic processes, especially by what I
have called homophony, the consonance of original and
mnemic excitations. Under such ontogenetic development
are included both the normal processes of growth and the
processes of regeneration and regulation.

This part of the book further undertakes to prove
that the above view of the relations between the Mneme
and the processes of genetics is not opposed to, but is
in accordance with, the results of modern experimental
research (Mendelism).

In this way I have tried to deduce from a common
property of all irritable organic substance — namely, that
of retaining revivable traces or engrams — a number of
mnemic laws, equally valid for the reproductions com-
monly grouped under memory, habit, or training, and
also for those which come under the head of ontogenetic
development, inherited periodicity, and regeneration
— laws common, in fact, to every kind of organic
reproduction.

I should like, however, to make it clear that while

14 THE MNEME

putting forward the mnetnic principle as the explanation of
a vast group of vital phenomena, I am far from imagining
that it explains the entire course of organic evolution
or the present state of the organic world. What I have
dealt with is merely the conserving principle necessary
for the maintenance of the alterations produced by
constantly changing environment. The adaptation of
the individual organism to the surrounding organic and
inorganic world is, of course, not explicable solely by this
purely conserving principle ; but, taking the latter
in conjunction with the principle of natural selection,
we shall, I think, find that the interplay of these two
great factors affords us an adequate insight into the
methods of organic evolution. The theory h.ere developed
will be found to be based solely upon causality, and to
require no help from either vitalism or teleology.

PART I

CHAPTER I

ON STIMULUS AND EXCITATION

My purpose in this book is to examine a specific kind of
stimulation, or more properly, of excitation. An exact
definition of the terms " stimulus " and " excitation "
is therefore an indispensable preliminary to any study
of the subject ; still, notwithstanding the brilliant results
already achieved by physiological research into stimuli and
their effects on animals, plants, and protists, a precise
meaning has not yet been assigned to the all-important
terms in question. In the first two German editions
of this work the greater part of the first chapter was
devoted to this task ; but having since treated it in a
separate essay, I will content myself in this chapter
with embodying its conclusions.

I will begin by stating the generally accepted definitions
of the terms employed. " Stimuli " are certain actions
on living organisms accompanied by specific effects. This
implies that we determine the nature of a stimulus by
the specific result thereby produced in an organism. It
is this result or effect which characterises the stimulus
as such.

What now is the specific nature of such action, as
distinguished from the nature of action not followed
by such result ? We may begin by advancing a negative
criterion : — " No action is a stimulus which does not
produce some corresponding physiological change." Thus
there can be no stimulus in the case of an inorganic
body, nor in that of an organism after extinction of
life. We are accustomed to define the changes referred

3 ''

i8 THE MNEME

to as a reaction of the living organism. Such reactions
can be regarded as falling into two main groups accord-
ing to the way in which we perceive them. One
group embraces the sensory reactions resulting from
impressions such as those of light, sound, pressure,
etc., given to us in immediate sensation. These im-
pressions can only be perceived by the individual himself
and the corresponding reactions are therefore described
as subjective. On the basis of extensive physiological
observations and experiments, we premise from these
respective sensations certain excitation - processes in
definite parts of an irritable substance.

The second main group consists of objectively perceptible
reactions, when an organism responds to a definite action
by a corresponding change perceptible by the observer, i.e.
a change which can be physico-chemically demonstrated.
This change may be a process of growth, or the contraction
of a muscle, or processes of metabolism such as secre-
tions or chemical redistributions. It is characteristic of
a great many of these objectively perceptible reactions
that they do not become manifest in that part of the
irritable substance subjected to the action of the stimulus,
but in a remote part of the organism. The most striking
examples of this are afforded by those tissues in which
the function of irritability has reached the highest degree
of specialisation, as in the nerve tissues of animals. But
examples may also be found amongst plants. As regards
nerve tissue, no immediate change can be noted either
in the exposed brain, or in the spinal cord, or in the nerves
originating in the same, when stimulated either electric-
ally, mechanically, chemically, or otherwise ; but accord-
ing to the part of the nervous system acted on, may be
observed the contraction now of this, now of that muscle
group, acceleration of breathing, change of heart action,
secretion of saliva and of tears. The result of a given
stimulus is to be observed not in the irritable substance —
the nerve tissue to which the stimulus was appUed, in
which no change is demonstrable either morphological
or chemical — but in possibly remote organs possessing

ON STIMULUS AND EXCITATION 19

the capacity of reaction to the given agent. We assume
with reason, however, that the primarily irritated substance
is affected, and it is now generally agreed to describe
this primary change as excitation, upon which the visible
reaction in the reactive organ follows only as a secondary
effect.

Long after this assumption had been generally accepted
as a clear explanation of the process involved, Du Bois
Reymond demonstrated in the electro-motive behaviour
of the nerves a state of excitement of the nervous substance
itself. So by means of the reversed polarity of the nerve
during stimulation (negative variation), we can prove
that the nerve substance is affected by the stimulus,
and we can in the same way directly demonstrate the
result of the stimulus on the irritable gland substance
by the negative variation of the gland current, and on
the vegetable cellular tissue by the negative variation
of the parenchyma current.

To sum up, if we group together all these specific
results of stimulation, the group will comprise certain
heterogeneous effects. First, the immediate sensations ;
secondly, effects observed on organs far remote from the
part directly subjected to the action ; and finally, effects
observed on the irritable substance of this part itself.

We assume, however, one common feature in this hetero-
geneous group, viz. the process of the excitation in the
irritable substance. The question as to the characteristic
effects of the action of stimuli may now be answered in this
way : — The effects of stimuli are known in all cases by
the appearance of an excitation in the irritable substance.

The existence of an excitation, as we have already
insisted, is purely a matter of inferential reasoning, and
the same reasoning points to excitation being some form
of energy ; for whether we base our reasoning on immediate
reactions in consciousness, or whether we argue indirectly
from our observation of motor or plastic reactions, meta-
bolisms, or the negative variation of the electric current,
we are obliged in all cases to assume an " energetic "
process in the irritable organic substance. It is impossible

20 THE MNEME

at present to state definitely what these energetic processes
are. Some writers believe that they are essentially
chemical energy. Others, in summary fashion, prefer to
speak of " physiological energy," or even of " nervous
energy," but they admit the possibihty of reducing this
into the well-known forms — mechanical, thermal, electrical,
radiating, and chemical — which may be termed elementary
energies. We prefer, however, to speak only of the ener-
getic process of the excitation, which may manifest itself
in many ways, differing according to the stimulus-receptor
which transforms the excitation.

Experience and experiment show that the entire,
highly complex state of excitement of the irritable
substance of an organism, which may be described as
its irritative-energetic condition, or, more briefly, as
its irritable condition, stands in close relationship to
the surrounding elementary-energetic condition which
is ceaselessly changing. We can, however, analyse this
relationship more precisely, for in many cases the specific
dependence of single components of the irritable condition
on single components of the elementary-energetic condition
can be traced.

This special relationship consists in the fact that the
initiation, duration, and termination of a component
of the elementary-energetic condition determines the
initiation, duration, and termination of a component of
the irritable condition.

The former component we term " stimulus," and the
latter " excitation." In ordinary speech we describe
such a relationship as one of cause and effect ; but in
all cases where we trace causal connection, that which
we call " effect " depends on a plurality of conditions.
There can be therefore no question of the production
of a single effect by a single independent factor. This
also obviously applies to the relation between the stimulus
and what has been defined as its " effect " — the single
excitation. In describing this relationship, therefore,
in terms of cause and effect, the presence of all other
essential conditions is implied.

ON STIMULUS AND EXCITATION 21

" Stimulus " may thus be defined as that elementary-
energetic condition whose initiation, duration, and termina-
tion in the presence of the requisite general conditions
are followed by, or, as we commonly say, " cause " the
initiation, duration, and termination of a single component
of the irritable condition or single excitation. This
component, the single original excitation, will thus not
only be initiated, but also maintained by the stimulus.
Further, a definite relation may be discerned between
the magnitude of the stimulus and the intensity of the
excitation resulting therefrom.

The above definition of the relation between stimulus
and excitation sets the time-relation in the foreground. It
describes the dependence, as regards time, of the excitation
on the elementary-energetic condition. For the stimulus
is immediately followed by the corresponding excitation,
which lasts while the stimulus continues, and which
ceases when the stimulus is withdrawn. For though
the excitation may not then immediately disappear, yet
it diminishes so rapidly that after a short interval no
traces of the same are to be found.

The time-relation between stimulus and excitation is
proved by observed and verifiable facts, and in my
judgment must be a chief factor in any close analytical
description of stimulation. I have already mentioned
that immediately on the cessation of the stimulus, not
complete disappearance, but rapid diminution of the
excitation takes place. The excitation exists in full
vigour only during the continuance of the stimulus,
appearing immediately after it, and rapidly subsiding on
the cessation of the stimulus. This, which we characterize
as " synchronous," is the main phase of stimulation, and
the excitation in this phase may be termed " synchronous
excitation."

On the cessation of the stimulus, the excitation rapidly
diminishes ; but always some seconds and sometimes
minutes elapse before the last demonstrable traces of
the excitation vanish, and the parts affected return to
the condition in which they were prior to the commence-

23 THE MNEME

ment of the stimulus. This phase of the diminishing
excitation, i.e. from the moment of the cessation of the
stimulus and the rapid fall of the excitation until its
total extinction, may be called the "acoluthic" phase
of the excitation. This phase also may be regarded as
a product of stimulation, or rather as a mediate product ;
the immediate product being the synchronous excitation,
upon which the acoluthic excitation follows. Regarding
the latter as a mediate product of the stimulus, we can
speak of an acoluthic or after-effect of the stimulus.

The study of acoluthic excitation has hitherto been
pursued very sporadically. Amongst sensory excitations
only visual acoluthic sensations have received detailed
attention. Some study has been given to auditory
acoluthic sensations ; but as regards the other fields of
sense, hardly any research work has been undertaken
so far.

In the department of peripheral, especially motor
excitations, those phenomena of acoluthic excitation
manifested on electrical stimulation of the muscles and
of the nerves have been very closely studied (opening
contraction, opening tetanus, etc.).

It must here be mentioned that the phenomenon of
stimulus-summation to be treated in the next chapter
implies a summation of acoluthic and synchronous
excitations, new synchronous excitations being added to
the already existing acoluthic excitations.

In the realm of plant-physiology, also, several cases
of " after-effects " have been observed. Unfortunately,
however, the acoluthic effect of stimulation has been
hardly, if at all, discriminated from the engraphic effect,
and the acoluthic and the mnemic excitations have in
most cases not been sufficiently distinguished. So far
as I am aware, only Sir Francis Darwin and Miss D. F. M.
Pertz have been alive to the fundamental distinction
between these two conceptions, and emphasised the
inaccuracy of applying the term " after-effect " indis-
criminately to both.
The acoluthic excitation is characterised by being the

ON STIMULUS AND EXCITATION 23

immediate continuation of the synchronous excitation,
which it terminates with a rapid diminution in intensity.

In many, perhaps in all cases, it possesses an oscillating
character, but the lowest point of the oscillations seems
to register not total absence, but merely a great weakening
of the respective acoluthic excitations, but in a few
seconds, or at most a few minutes after cessation of
the stimulus, the acoluthic excitation entirely vanishes,
and no further traces of it as such, i.e. as a demon-
strable excitation, can be discerned. The condition of
the irritable substance as regards the special excitation
which has now died away is apparently the same as before
the application of the stimulus. This state of renewed
indifference may be called the secondary state of indiffer-
ence, in contrast to the primary state of indifference
which existed prior to the application of the stimulus.
The ordinary understanding is that primary and secondary
states of indifference are practically, if not absolutely,
identical. They are so in respect of the immediately
manifested reactions, but it will now be our task to show
that they are not so in respect of their capacity for reaction.
Plant-physiologists have given more attention to this
matter than animal-physiologists, but both have some-
what neglected the systematic handling of the problem.
The view here taken is that the problem is of funda-
mental importance in the investigation of the physiology
of stimulus and for the theory of descent.

CHAPTER II

ENGRAPHIC ACTION OF STIMULI ON THE
INDIVIDUAL

When an organism has been temporarily stimulated
and has passed, after the cessation of the stimulus, into
the condition of " secondary indifference," it can be
shown that such organism — be it plant, protist, or animal
— has been permanently affected. This I call the engraphic
action of a stimulus, because a permanent record has been
written or engraved on the irritable substance. I use the
word engram to denote this permanent change wrought
by a stimulus ; the sum of such engrams in an organism
may be called its " engram-store," among which we
must distinguish inherited from acquired engrams. The
phenomena resulting from the existence of one or more
engrams in an organism I describe as mnemic phenomena.
The totality of the mnemic potentialities of an organism
is its " Mneme."

In adopting original expressions for the conceptions
defined in this book, I am not unmindful of the words
" Memory " and " Memory image," already in use.
But for my purpose I should have to use these in a much
wider sense than is customary, and the way would thus
be opened for innumerable misconceptions and consequent
controversy. It would be a mistake to give a wider
interpretation to a term which is ordinarily used in a
narrower sense, or to use an expression like " Memory
image," which invariably suggests phenomena in
consciousness.

In quoting experimental examples of engraphic stimu-

ENGRAPHIC ACTION ON THE INDIVIDUAL 25

lation in higher and lower organisms, one thing must
be borne in mind. The abiUty to retain the engraphic
stimulations — the so-called engraphic susceptibility —
varies in irritable organic substances, just as the irrita-
bility in respect of synchronous excitation differs greatly
in various organisms, and in the varieties of tissue and
cell within the same organism. In animals, during the
evolutionary process, one organic system — the nervous
system — has become specialised for the reception and
transmission of stimuli. No monopoly of this function
by the nervous system, however, rail be deduced from
this specialisation, not even in its highest state of
evolution, as in Man. To quote a significant case, it
has been demonstrated by indisputable observation and
experiment that muscle is still irritable even after entire
elimination of all nervous influence.

Just as the synchronous irritability of the nervous
system has gradually increased in the evolution of the
species, so its engraphic susceptibility has increased.
Yet neither of them has become a monopoly of the nervous
system, but has remained in the higher organisms as a
property of irritable substance as such, thus seeming to
be indissolubly bound up with the mere quality of irrita-
bility. Observations on the nervous system lead us to
the conclusion that the engraphic susceptibility grows
with the increase of irritability. Weak and momentary
stimuli may fail in any decisive engraphic effect on the
non-nervously differentiated organic substance, and yet
exercise a strong influence so far as nerve substance is
concerned.

This idea is advanced in order to prepare the reader
for the demonstration of the fact that the engraphic
action of stimuli on nerve substance is simpler and more
direct than on non-nervously differentiated substance,
especially in cases of the experimental production of such
engraphic action. On non-nervous substance, as con-
trasted with nerve substance, the stimuli have as a rule
to act much longer, or repeat themselves much more
frequently, in order to produce engraphic effects. But

26 THE MNEME

in the nervous substance of the higher animals a single
momentary stimulus suffices to produce an easily demon-
strable, durable engram. In principle httle significance
attaches to this difference, but it is strikingly evident
in experimental illustrations, which gain in effectiveness
and simplicity in proportion to the amount of nervous
differentiation in the material employed.

For this reason we may take as our first example of
engraphic stimulation one on the nerve substance of a
higher animal. The supposition that we are best able
to study the physiological properties and capacities of
organic substance by experiments with unicellular living
creatures is an error, though common enough in the
scientific literature of the present day. Where the
division of labour amongst cells and tissues is far advanced,
and specific organic functions have been evolved, our
study of these functions is made more simple and the
results of our work are less ambiguous than in those
cases where functions are merged in each other, and it
seems almost impossible to differentiate them.

Although in our study of engraphic stimulation both
nerve and other substance must be taken into account,
it will be easier and better to proceed in our research
from the more to the less differentiated.

Let us take first the case of a young dog whose simple
trust in humanity has not yet been shaken. He is pelted
with stones by boys at play. Two groups of stimuli act
upon the creature : — (Group " a ") the optical stimulus of
the boys picking up stones and throwing them, and
(Group "b") the tactual stimulus of the stones striking
the skin, resulting in pain. Both stimuli groups act
engraphically, for after the cessation of the synchronous
and the acoluthic effects of stimulation, the organism
now appears permanently changed in relation to certain
stimuli. Previously, the optical stimulus of a stooping
boy was accompanied by no definite reaction, but now
this stimulus acts regularly — generally until the death
of the animal — as a pain-causing stimulus. The animal
puts its tail between its legs and runs away, often with

ENGRAPHIC ACTION ON THE INDIVIDUAL 27

loud howls. We may express this by saying that the
reactions belonging to stimulus group " b " are henceforth
elicited not only by these stimuli, but also by stimulus
group " a."

The illustration affords a deeper understanding of
the nature of the engraphic stimulation. We perceive
that the change of the organic substance is of such a
nature that the synchronous state of excitement belonging
to stimulus " b " may be aroused not only by the entry
of stimulus " h" as in the primary state of excitement,
but also by other influences, such as the stimulus group
" a " in the above-mentioned case. The influences
which produce this result may be termed " ecphoric "
influences, and if they possess a stimulus character,
ecphoric stimuli. But, as we shall find later, not all
ecphoric influences can simply be called " stimuli."

It is only when the synchronous and acoluthic action
of a stimulus has ceased, and the substance has entered
the secondary state of indifference, that we are able
to recognise whether the effect left behind is an engraphic
change, and this we are able to do by the following
method: — We have to find out whether the state of
excitement belonging to that stimulus may now be
called up by influences differing quantitatively and
qualitatively from the engraphically acting stimulus. A
stimulus which on its first appearance acts engraphically
may be called an original stimulus. The synchronous
excitation with its acoluthic sequence may be termed
an original excitation.

It is self-evident that the stimulus produces at each
action its corresponding synchronous state of excitement,
and that this is in itself no proof of a preceding engraphic
change. A stimulus must therefore differ either
quantitatively or qualitatively from the original stimulus
before one can say of it, from the standpoint of objective
analysis, that it acts ecphorically. To prove this fully,
it is necessary to demonstrate by experiment that this
ecphoric influence is either quantitatively or qualitatively
insufficient by itself to elicit the given reaction without

28 THE MNEME

the preceding action of the original stimulus. In the
case before us it is easy to obtain proof by comparing
the behaviour of the animal before the painful experience
with its behaviour after it. The essential points of the
case quoted, to which many other cases of animals, birds,
reptiles, and cephalopods might be added, can be stated
in the following propositions : —

(i) Stimulus " a," as original stimulus, generates only
excitation "a."

(2) Stimulus " b," as original stimulus, generates only
excitation "j3."

(3) Excitation (a + j8), as original excitation, is
generated only by stimulus {"a" + "b").

(4) But excitation (a + j8), as mnemic excitation, that
is, after former action of stimulus (" a " + " b ") and the
creation of the engrams (A + B), may be generated by
stimulus " a " alone, acting as ecphoric stimulus.

In the case of lower animals, plants, and protists, it
is as a rule impossible to obtain an engraphic effect by
the action of a single momentary stimulus. A prolonged
or frequently repeated stimulation is required. But we
are still of opinion that further research will furnish
convincing experimental evidence' of engraphic effects
of all kinds in the case of the lower organisms.

Proof of clear and permanent engraphic effects may
be gathered from observations made during the last
decade by Bohn, Jennings, van der Ghinst, and others
on non-vertebrates, such as the lower Molluscs, Echino-
derms, Coelenterates, and the more highly organised
Crustaceans. Hodge and Aikins as long ago as 1895
noted such effects, traces of which remained for several
hours.

Since the first edition of The Mneme Sir Francis
Darwin, pursuing the line of investigation suggested in
that book, has collected various examples of united
' (associated) engrams in the realm of the physiology of
plants. Even earlier, the same distinguished worker
in collaboration with Miss D. Pertz succeeded in associating

ENGRAPHIC ACTION ON THE INDIVIDUAL 29

the photical or geotropical engrams in plants with metabolic
engrams.

Engraphic effect can readily be demonstrated by the
application of one single kind of stimulus. For after
repeated or prolonged action of the stimulus, and after
return of the organism to the secondary state of indifference,
a much feebler stimulus of the kind which produced the
original state of excitement suffices to produce the same
effect. Such effects may be produced in animals with
moderately high nervous systems by a few momentary
stimuli. Thus Davenport and Cannon, in the course
of experiments made for quite a different purpose on
Daphniae, discovered that their reaction to the stimulus
of light — a positive heliotropic reaction — altered per-
ceptibly on the application of a limited number of
momentary stimuli. After this application only a quarter
of the original light-stimulus was required to produce
the original or an even stronger reaction. This result
was a constant one. Observations frequently made by
botanists and students of the Protista on the alteration
of the so-called " light disposition," under the influence
of photic stimuli, fall into the same class. The reactions
by which these alterations, surviving the supervention
of the secondary state of indifference, become manifest
may be motor reactions as well as those of growth. In
connection with the latter, Oltmanns, experimenting
on fungi, exposed them for ten hours to the electric light.
For a further fifteen hours the plants were kept in the
dark, and then again subjected to intense light. " Under
the influence of the renewed exposure the sporangia at
first made some strongly negative curvatures, which
were soon neutralised. Then within a relatively short
interval positive movements set in, which were maintained
with far greater energy than previously, and which also
brought about sharper endings. It is clear that the
latter phenomenon was caused by the preceding exposure
to intense light, that is, that the movements became
more energetic in consequence of increased " light
disposition."

30 THE MNEME

The essential points of this and the previous observation
may be expressed in propositions corresponding to those
in which we summed up the results of our first example
(p. 28).

(i) Stimulus "a." as original stimulus, generates

original excitation "a."

" a "
(2) Stimulus , as original stimulus, generates original

till 2
a

excitation only.

^ " a "

(3) But stimulus , as ecphoric stimulus, may

liberate excitation " a " as mnemic excitation (i.e. after
the previous action of stimulus " a " and the creation
of an engram " A ").

Further proof of the engraphic effect of the stimuli in
respect of the excitations, both in the individual subjected
to them and in his progeny, will be given later. For
the present it will be more to our purpose if, instead of
quoting further cases, we begin the specific analysis
of engraphic stimulation and its manifestations, an
analysis, however, which is merely provisional at this
stage, and will be more fully elaborated in the second
part of the book.

We may begin by considering the organism in its primary
state of indifference.

Primary State of Indifference. — Under this defini-
tion we have simply to conceive the state of the organism
at the beginning of the observations and experiments.
Into this state it is absolutely essential for us to enquire
minutely. At the outset we are met by difiiculties of
a two-fold order ; for the objects which we choose for
analysis possess — unless they are germs just separated
from the parent organism — a sum of individually acquired
engrams, and also engrams resembling those whose genesis
we wish to observe, or which we wish to produce artificially.
Suppose we choose a one-year-old specimen of the sensi-
tive plant Mimosa pudica, and try to influence it

ENGRAPHIC^ ACTION ON THE INDIVIDUAL 31

engraphically by photic stimuli, it will not be sufficient
to register at the beginning of the experiments its reactions
to light during the previous twenty-four hours. The
reactions given at the end of September by a one-year-
old plant grown in Christiania may be almost identical
with the reactions given by a plant imported direct from
the Equator, but we must not forget the possibility of
the manifestation to pronounced differences a few months
later, which would be inexplicable did we not take into
account the engrams which the plants acquired in their
different places of origin. One way out of this difficulty
would be to cultivate the demonstration objects, whenever
practicable, directly from seeds and ova, and keep them
under conditions which safeguard a study of the engraphic
action of such stimuli. A mimosa, which for one year had
been subjected alternately to twelve hours' artificial
light and twelve hours' artificial darkness in a room kept
at an equal temperature, is a much more controllable
object than one which has been exposed only to natural
conditions of light and darkness. The best results, how-
ever, are obtained from those objects upon which, in
their individual lives, relatively few stimuli have acted,
e.g. the seed plant just emerging and exposed for the
first time to the light of day, or the young chick at the
moment of breaking its shell. Whatever object is chosen,
the attempt should be made to secure individuals which
hitherto have not been subjected to that particular
stimulus, of which it is intended to study the engraphic
action.

So far, our method of determining engraphic change
by simple observation with a minimum of inferential
reasoning, has consisted in demonstrating the change of
the capacity to react between the primary and the
secondary states of indifference. The fewer individually
acquired engrams that exist in the primary state, the less
complicated the task.

The individual, which in its unicellular stage has just
separated from the parental organism, is virgin soil as
far as its individual Mneme is concerned ; but as we shall

32 THE MNEME

see later, it already possesses inherited engrams, indeed a
considerable stock of them. Accordingly, since on our
planet it is impossible to obtain fresh, spontaneously
generated organic material, there is no single organism
available for examination which, mnemically considered,
can be regarded as tabula rasa. The germ cell, which
a moment before was part of the parent and shared its
Mneme, does not by the act of separation from its parent
and by its entrance into a new individual phase obliterate
entirely its own Mneme. This might, of course, be
assumed a priori, but it can be demonstrated by an
abundance of evidence.

Later on it will be our task to see to what degree the
germ cells share in the individual and acquired engrams
of the entire organism, and to what extent they retain
their share after separation.

Engraphically Acting Stimulus. — Energetic influ-
ences from those energy-groups, of which we know that
they produce synchronous excitations in organisms, may act
engraphically simply by the medium of those excitations.
Such are mechanical, geotropic, acoustic, photic, thermal,
electric, and chemical influences. Mag netic influences
appear to be akogej^ier,j!mail§__jtQ.-act„.a_s_ stimuli - on
organisms. If this assumption be correct, then neither
can SQtlrinfluences act engraphically. Again, it is possible
that energies which have hitherto escaped our observation
may produce synchronous excitations in organisms, and
thereby also act engraphically. Within recent years,
the " X " rays — a hitherto unknown kind of radiant
energy — have been discovered. Shortly afterwards, their
capability to act on organic bodies as stimuli was demon-
strated. Of Radium radiation the same may be said.
The above enumeration does not exhaust the list of
energies which act as original stimuli, and in a secondary
degree as engraphic stimuli. It suffices, however, for the
present phase of our research.

The question now arises : — When does a stimulus that is
capable of originating an engraphic effect act engraphically,
and when does it not ? To answer this question we must

ENGRAPHIC ACTION ON THE INDIVIDUAL 33

examine some general laws of stimulation. To produce a
synchronous effect each energetic action must be of a
definite strength and duration, and these vary in value
according to the nature and state of the organism to
be influenced. In this sense, we may speak of the
" threshold " of the stimulus. Careful observation shows
that this " threshold " or liminal value depends not
only on the strength and duration of the energetic action,
but also on a third factor, namely, its continuity or its
discontinuity. Whilst the first two factors in their
significance for synchronous stimulation do not require
further comment, the third factor demands more minute
treatment. It is well known that electrical (as also
mechanical) actions on contractile substances, l3nng
usually under the threshold, become effective after repeated
action. Evidently, by the continued application of
stimuli, the irritability of the organic substance is
increased by " addition latente " (Richet), which lowers
the threshold of the effective intensity of such stimulus
for the given substance, so that a hitherto subliminal
intensity becomes a liminal one. Biedermann, demonstrat-
ing the stimulus-summation of non-striated muscles,
writes : — " Even under the most favourable conditions
a visible effect (contraction) can hardly be obtained by
strong, but detached induction shocks ; whilst the same
objects, namely, the muscles of Molluscae — all of which
are non-striated — and the muscles of the intestines and
of the ureter, manifest tetanus with even a relatively
weak current, when the stimulus acts in quick succession
under Neef 's oscillating hammer. Even in the application
of constant currents, according to Engelmann, one can
often induce an effective excitation by frequently making
and breaking a current which otherwise would not act.
The capacity of stimulus-summation, although graduated ^
according to development, seems to belong to all irritable
plasm — to ciliated cells, to nerve cells, and to vegetable
plasm such as that of Dionsea, etc. — ^so that the recorded
manifestations in the above-mentioned experiments on
muscles represent only a specific case of a general law."

3

34 ' THE MNEME

Recent investigations by Steinach regarding the flagellates,
infusoria, and plant and animal tissues, have fully con-
firmed the accuracy of Biedermann's view that all irritable
plasm is susc^tible of stimulus-summation. We may
instance the very great capacity for summation of vegetable
cells and the still greater capacity of the luminous cells
of Lampyris.

In considering the engraphic action of discontinuous
stimuli, we shall have to differentiate two distinct classes
of effects. It is apparent that no stimulus acting en-
graphically does so directly, but only through the medium
of the excitation produced by it. If a discontinuous
stimulus produces a continuous excitation which occurs
regularly under short intervals of stimulation — as in
the cases described by Biedermann and Steinach — the
excitation so produced does not differ in its engraphic
action from that produced by continuous stimulation.
The former may be stronger quantitatively as a result
of stimulus-summation, but is otherwise identical.

The engraphic action is, however, altogether different
if the repetition of the stimulus takes place at longer
intervals of time, and if the periods of stimulation are
sufficiently separated for the excitation produced by
the preceding stimulus to have died away before a re-
application, and for the irritable substance to have
returned to its corresponding state of indifference.
In such cases, not only the stimuli, but also the excitation
produced by them, are discontinuous, and each repetition
of the stimulus then produces a fresh engram qualitatively
equal to, but distinct from, its corresponding predecessors.
This discontinuity and independence of engrams which
owe their existence to the repetition of the same stimulus
is of great importance, as will be explained in detail in
Chapter VII on the subject of Homophony and the signifi-
cance of stimulus-repetition.

So far we have spoken only of the engraphic action of
a single continuous or discontinuous stimulus. Each
organism, however, is affected by conditions of distance
and volume and motion in regard to the various kinds of

ENGRAPHIC ACTION ON THE INDIVIDUAL 35

energy to which it is permanently exposed. We defined
these conditions as its external energetic situation.

Now, it appears hardly possible, even in the best planned
experiment in a laboratory, to alter the eili^getic situation
in relation to a single energy alone, and we can hardly
expect to meet with such a case under natural conditions.
When the sun breaks through the clouds and shines
on a plant, not a simple, but a highly complex change
of the energetic situation is produced, and different kinds
of radiant energy, such as ultra-red heat rays, various
light rays, and chemically acting ultra-violet rays, act
as so many stimuli on the organism.

Only in the laboratory are we able in a measure to

allow purely photic influences to act on the organism —

for instance, by applying red rays of a definite wave

length which exercise hardly any chemical effect, and

the thermal influence of which has been reduced almost

to zero by passing them through interpolated films of

ice. A mimosa standing in a darkened room may be

subjected suddenly to the influence of sunlight. By this

single cause is produced the simultaneous action of at

least three stimuli, the synchronic effect of which we

can demonstrate by the appearance of three different

reactions. To the photic stimulus the plant responds

by unfolding its leaves, to the chemical stimulus by

metabolic reactions known as the absorption of carbonic

acid and the emission of oxygen, and to the thermal

stimulus by acceleration of growth. By eliminating

either the thermal, photic or chemical rays, we can

demonstrate how the previous simple change of the

external energetic situation was the occasion of different

simultaneous stimuli. The energetic situation of the

organisms on our planet is continually subject to change,

not in one, but in many, respects. Such changes may have

a manifest connection with each other. For example, a

thunderstorm produces simultaneously photic, thermal,

acoustic, mechanical and numerous other stimuli affecting

those organisms with whose energetic situation it interferes.

Often various stimuli simultaneously influence the same

36 THE MJSIEME

organism, and we are unable to trace their relations of
origin. Such a coincidence we describe as fortuitous.

We conclude that each organism is continually subjected
to stimulation, and in most cases to the simultaneous
action of various stimuli.

We may now ask whether two or more stimuli, acting
simultaneously on an organism and producing synchronous
effects, can also influence it engraphically in juxtaposition.

The answer is readily given in the case of those organisms
where the engraphic effects of stimuli are easily observed.
If the organism be a man, a monkey, a dog, a horse, or
a bird, it is comparatively easy to demonstrate the
engraphic action of two simultaneous stimuli. If I whip
a young dog which has never been punished before,
both the optical stimulus — the sight of the whip — and
the tactual stimulus — producing sensations of pain — act
engraphically and, what is of particular importance,
the engrams produced by these simultaneous stimuli
enter into a certain indissoluble relationship to each
other. This relationship can be defined by stating that
henceforward the application of only one stimulus will
suffice to act ecphorically on the engram simultaneously
produced by the other stimulus. The mere sight of the
whip in the hand of its master will ecphorise in the dog
the mnemic excitation of the definite sensation of pain,
and produce the corresponding reaction, namely, curling
of the tail, howls, and flight. We describe such engrams,
where the application of the engraphic stimulus of one
serves as the ecphoric stimulus of the other, as associated
engrams. We may state it as a rule without exception
that all simultaneously-produced engrams are associated
even when the effective stimuli are of different kinds
and have no relationship in respect of the cause of their
appearance. Two stimuli of very different kinds and
without recognisable relationship once acted upon me
simultaneously — the view of Capri from Naples, and a
specific smell of boiling olive oil. Ever since, a similar
smell of oil unfailingly acts ecphorically on the photogenic
engram of Capri,

ENGRAPHIC ACTION ON THE INDIVIDUAL 37

Apart from this association of engrams created by the
simultaneous action of their engraphic stimuli, which,
in keeping with the terms already introduced, we may
describe as association of simultaneously created engrams,
a second and equally important association may be
observed, which is also dependent on the time-relation
of the action of the engraphic stimuli. Not only the
simultaneous engrams, but also those generated immedi-
ately after, are associated in such a manner that the
application of the original stimulus of one may serve as
an ecphoric stimulus to the others. In this case, too,
an association is established, even when the respective
engrams owe their genesis to widely differing kinds of
stimulus, and where a causal nexus between these engraphic
stimuli cannot be traced. We may describe this latter
condition as the association of successively created engrams.
We shall examine more closely in the second part of
this work the derivation of the successive from the
simultaneous association.

Secondary State of Indifference. (Latent state
of the engram.) — After the cessation of the synchronous
state of excitement, followed by a transient acoluthic
state of excitement, the organism passes into a state
which may be .termed the secondary state of indifference.
This is distinguished from the primary state solely by
the possession of the new engrams due to the action
of the various engraphic stimuU. These, however, being
latent are not perceptible in the secondary state.
Ecphoric influences are necessary for their manifestation.
The intervention of a latent phase between the
synchronous and the mnemic states of excitement is
highly characteristic of mnemic phenomena, and presents
the mnemic excitation in the Ught of a reproduction.

The results of our researches so far may be summarised
as follows : — A stimulus induces in an organism a specific
state of excitement which manifests itself by definite
reactions. On the cessation of the stimulus, the state
of excitement is at once modified, and a little later alto-
gether subsides. The irritable substance, so far as the

38 THE MNEME

stimulus is concerned, returns to the state in which it
was before the action of the stimulus, and the return
marks the secondary state of indifference. But the
two states of indifference, the one before and the other
after the interference of the stimulus, are not identical.
They differ in that the irritable substance in its secondary
state may, by the action of certain ecphoric influences,
manifest the intermediary state of excitement, a result
which would not be possible in its primary state. The
question arises whether there are not cases where the
irritable substance, after the cessation of stimulation,
does not pass into the secondary state of indifference,
but maintains permanently the state of excitement
originally created by the stimulus. I believe we are
to-day already in a position to answer this question
with a decided negative, notwithstanding that our re-
searches are not yet completed. Of course, there are
subsequent results of the excitations which do not dis-
appear with the cessation of the excitation, for instance,
the results which appear as phenomena of growth.

More difficult to judge are those cases which Pfeffer
(Plant Physiology, vol. ii, p. 167) describes as continuous
(static, inherent) induction. In these cases, a temporary
stimulus creates a continuous induction. May this not
be explained on the assumption that here also the
stimulation results in growth, which itself conditions
new growth by acting as stimulus to that end ? In
none of these cases can we speak of persistence of the
irritable substance in its primary state of excitement.
The case of Marchantia polymorpha seems to present
considerable difiiculty. The gemmae of this liverwort
can be so influenced by a few days' exposure to light
from one side, that it becomes definitely determined
which is going to be the upper and which the lower side
of the plant before the anatomical differentiation has
become pronounced in the seedling.

It may be, however, that future research will show
that the action of stimuli creates permanent states in the
morphological structure of the growing seedling, and that

ENGRAPHIC ACTION ON THE INDIVIDUAL 39

these states, acting as stimuli of position, continue to
influence new acquisitions. But until such a structure
has been demonstrated, this interpretation of the difficulty
must be regarded as merely an assumption. Further,
since other explanations are also conceivable, the case
of Marchantia need not be taken as proof that under
certain circumstances a state of excitement becomes
permanent in the organic substance, instead of ultimately
resolving itself into the secondary state of indifference.
The general order of things is as follows, and may be
stated as an established rule : — After the cessation of
a stimulus the organism returns sooner or later into the
state of indifference. A permanent effect on the irritable
substance is produced by the stimulus only in so far
as it leaves an engram behind. The irritable substance
is therefore altered in this respect that, from now, the
synchronous state of excitement peculiar to the stimulus
may be resuscitated not only by the latter, but also by
other influences, which may be termed ecphoric.

Ecphoric Influences. — From the preceding exposition
it follows that the engram of a stimulus — that is, the
engram of the state of excitement produced by a stimulus —
is simply the altered disposition of the irritable substance
in its relation to the repetition of the state of excitement.
The organic substance reveals the working of a new law,
in that it is now predisposed to the state of excitement
by influences other than that of the original stimulus.

The state of excitement arising from the ecphorizing
of an engram may be described as the mnemic state of

excitement. And so far as a mnemic excitation manifests

itself as a sensation, the latter may be termed a mnemic

sensation.
Surveying the mnemic phenomena in the three organic

realms, we find that the following groups of influences may

act ecphorically on an engram : —

First : — ^The repetition of the original stimulus either '

in a qualitatively and quantitatively identical — or almost

identical — ^form, or in a form which, while similar, varies

somewhat quantitatively and quaUtatively.

40 THE MNEME

Second : — The ecphory of other engrams generated
at the same time as the engram, or immediately
before. Engrams may be associated simultaneously or
in succession.

Third : — Certain influences which appear to us at first
only as definite periods of time or development also act
ecphorically.

These may all be reduced, as we shall see later on, to
the one princ iple which underlies all three groups. This
principle is the partial recurrence ot a aennite energetic
condition.

The simplest of all quoted cases is apparently that in
which a stimulus, identical qualitatively and quantita-
tively with the original stimulus, acts ecphorically. But
in this particular case it is hardly possible to prove the
ecphoric action of the stimulus objectively. For one has
no right to difEerentiate between the effect of a stimulus
at its first and at its later action if, on its recurrence, it
shows exactly the same phenomena as on its first appear-
ance. Nevertheless good reasons may be given for
speaking of an ecphoric effect even in the case of a mere
repetition of the original stimulus. These reasons are
based in the first place on subjective observation or
introspection. This essential aid is regarded dubiously by
many scientific men, whose work may not have allowed
them to become intimate with the methods of physiological
and especially sense-physiological research. But as we
shall often have occasion to refer to introspection, some
remarks on its value as a method for scientific research
may well be made.

AH that we name the " external world " is cognisable
by us objectively, that is, by occurrences and processes
within our own organism. A number of retinal and
brain processes result in the consciousness of what we
call a " tree," which we regard as external to us. Other
processes involving the olfactory epithelium we describe
as " fragrances," referring them also to something external.
Consciousness involves the perception of stimulus effects.
We gradually learn to argue from these effects to the

ENGRAPHIC ACTION ON THE INDIVIDUAL 41

stimuli causing them. Experience helps us to differentiate
" inner " and " outer.'-' The developing child thus builds
up for itself an image of the outer world which, like the
unsophisticated adult, it refers to something not itself.
It is altogether unaware of the subjective foundation
of its external world. At no time must we lose sight
of the fact that subjective states of consciousness are
for us the prime foundation of the externeil world. From
them we fashion the world of images which represents
for us external reality, and which, by the use of analogy,
we come to regard as objective. So much at least must
be acknowledged, even if for the purposes of this book
— as for science — the reality of this external world is
unquestioned. The necessity of employing the subjective
(introspective) method in the analysis of many funda-
mental problems of biology and physiology hes in its
being the only one whereby knowledge based on sense-
perception can be obtained. By the objective method
the analysis of sense-perception is impossible. Even
when to the methods of observation and experiment we
add the method of verbal communication, the ideas
we form by such objective means of the sensations of
our fellow creatures are not particularly well defined,
and probably only quite exceptionally, if at any time,
absolutely correct. The application of the subjective
method to the problems of biological science is, therefore,
perfectly legitimate. It has been applied in proper places
by physiologists such as Johannes MiiUer, Helmholtz,
Hering, and many others, men of the highest repute in
their profession. We owe the most valuable results
of sense-physiology to this method, without which the
imposing and securely founded edifice of this science would
never have been built. But we need to be on our guard
against a non-critical intermixture of the objective with
the subjective method.

The latter method is purely personal, but the use of
the objective method gives material for comparison,
whereby we can check the observations of our own
stimulations by reference to the experience of others.

42 THE MNEME

But as soon as we begin to apply conclusions from the
observation of reactions occurring in our own muscles,
glands, etc., to the sensations of other individuals we at
once enter the field of hypothesis.

This hypothesis is perfectly justifiable as long as we
are concerned with those creatures which closely re-
semble us morphologically and physiologically, and whose
physiological reactions speak to us in a language hardly
to be misunderstood, especially when human speech
is available as an auxiliary. Only I myself know the
qualitative and quantitative values of the sensation of
pain caused in me by the prick of a needle, and can
estimate them to a fairly correct degree ; but it would
be hypercritical to reject the conclusion that this stimulus
causes a sensation nearly identical with that in my fellow
men who react in a similar manner by a twitch and a
sudden cry, and who can minutely describe to me in
words the sensation they feel. It is very probable that
a monkey or a dog, reacting against a needle-prick by a
twitch and a specific vocal sound, feels something similar
to my own sensation. We are, however, on less certain
ground when we consider the reactions that a frog or
a fish may make to such a stimulus as a needle-prick,
although it is probable that the sensations are not unlike
our own.

Our conclusions become still less valid when we approach
the invertebrate Worms, the Coelenterates, and similar
creatures, even though these react against the pin-prick
by a sudden withdrawal of the part touched. But an
assertion that a Mimosa feels something Uke our sensation
of pain at a touch, against which it reacts by the sudden
folding up of its leaves, falls entirely outside the limits
of scientific discussion because of the lack of those con-
vincing inferences from analogy which might make it
appear probable.

The subjective method of the investigation of Self
gives results which are denied in great measure to the
objective method. We may apply the former in unlimited
measure in our perceptive knowledge of nature. It

ENGRAPHIC ACTION ON THE INDIVIDUAL 43

is legitimate, moreover, to make analogical use of such
knowledge in our estimate of the sensory reactions of
other creatures. Human intercourse is largely dependent
on such a use. But we must bear in mind the uncertainty
of the conclusions based on analogical reasoning, for
the validity of them diminishes as we extend them from
our fellow men to the remaining warm-blooded species,
from these to the cold-blooded ones, from the Vertebrates
to the Non-vertebrates, and from these to the Protozoa
and Plants. In the course of our present investigation
we are fortunately able to dispense with such a transference
of the subjective method. We apply this method only
to the Self, so that each reader may form for himself
a clear idea of the ecphory of an engram by a stimulus
resembling quaUtatively and quantitatively the original
stimulus. Everybbdy may observe in himself, on the
repetition of a stimulus which has already once acted
upon him — for instance, the optical stimulus of an
arabesque or the characteristic design of a carpet or
wall-paper — that the state of sensation which arose at
the first action of this stimulus is not simply repeated,
but that a new element appears in consciousness, namely,
the sensation of having been subjected once before to
this specific stimulation, that is, of having already ex-
perienced this characteristic state of excitement. We
describe this state of consciousness which responds to
stimuli of all kinds, and which we do not intend to analyse
here any further, as " regognitiaa." The point we wish
to emphasise is that the irritable substance in the secondary
state of indifference, as compared with it in its primary
state of indifference, has undergone a change, or, as we
say, has been influenced engraphically. If this repetition
of the original stimulus produces the same synchronous
effect as at its first appearance, and if my consciousness
tells me clearly that my organism has retained traces
of having already passed through this state of excitement,
then here surely is first-hand evidence that the second
stimulus does not only exercise a synchronous effect,
but also an ecphoric one in making me aware of the

44 THE MNEME

existence of an engram by means of a specific reaction,
namely, that of consciousness.

With the assistance of the objective method it is
possible to show in a fairly convincing but more uncertain
way that the existence of an engram may be traced
by the simple repetition of the original stimulus. This
less direct proof is based on the fact that the objectively
traceable reactions frequently appear in quicker or stronger
fashion on repetition of the stimulus than when the
stimulus was first applied. We may remind the reader
of the previously-quoted observations of Davenport and
Cannon, which show that Daphnise in their heliotropic
movements towards the source of light require at the
third application of a strong light-stimulus about half
the necessary time (twenty-eight instead of forty-eight
seconds) to swim the distance of i6 cm. The investigations
of Oltmann into fungi lead to similar conclusions. The
experiments cited lead to cases where a stimulus coinciding
qualitatively with the original stimulus, but differing
from it quantitatively, can be regarded as an ecphorising
stimulus. After repeated stimulation of an organism,
a stimulus of far less intensity and duration than was
required in the first instance will suffice to produce the
same reactions (lowering of stimulation threshold).
Evidence of this being an ecphory of engrams, however,
is given only by those cases where the repetition does
not take place at too short intervals, so that the organism
has time to pass entirely into its second state of indifference
and to allow of the subsidence of the synchronous and
acoluthic states. It is not necessary, therefore, " for us
to consider the phenomenon of stimulus-summation,
for, as already suggested on page 34, such summation
is not a case of the ecphory of mnemic excitations by
the rapid repetition of stimuli, but rather a case of
synchronous excitations freshly generated by these stimuli,
added to acoluthic excitations which have not yet subsided.
We may see engraphic and ecphoric action in the experi-
ments by Davenport and Cannon on the Daphnise,
where, after repeated stimulation at gradually longer

ENGRAPHIC ACTION ON THE INDIVIDUAL 45

intervals, a stimulus possessing but a quarter of the
strength of the original stimulus produced after a few
repetitions the same reactions as the original stimulus,
and after many repetitions even quicker reactions than
were given by the original stimulus on its first application.
Similar conclusions may be obtained from observations
made on the higher animals and on man. In the breaking-
in of a horse, e.g. while training it to canter, the trainer
may reduce the force of the pressure-stimuli, which induce
definite positions and movements of the horse, so much
that comparatively weak stimuli which at the beginning
were not even noticed will at the end be quite effective.
By the repeated action of stimuli it is possible, therefore,
to lower their stimulation threshold. Most frequently,
these are instances of ecphoric action on engrams ; and,
exceptionally, of the general increase of sensitiveness
in the respective sensory areas. A good wine-taster is
not necessarily also a good tea-taster ; and although
perhaps better qualified to be a taste expert than an
ordinary mortal, he first must acquire hundreds of new
engrams from the chemical stimulus of tea in order
to be able to react to the finely-graded differences of
stimuli which would be bewildering to the non-expert.
Similar examples might be adduced in reference to the
engraphic training of the visual and oral senses, but in
these cases it is more difficult to exclude from these
reactions the co-operation of subsidiary brain-processes
than from those of the senses of touch, taste, and smell,
which are less complicated with accessory processes.
Slight qualitative deviations from the original stimulus
do not, however, invalidate the ecphoric action of a
stimulus on the engram belonging to the original stimulus.
Thus, it is sufficient to see the picture of a landscape
in order to be able to ecphorise the engram belonging
to the landscape itself ; to hear a tune hummed in order
to ecphorise the engram belonging to the original full
orchestral performance. If the smell of Selene gas
ecphorises in us the olfactory engram of rotten radish,
it is clear that one ecphoric stimulus has vicariously

46 THE MNEME

replaced another which from a chemical point of view
is altogether different. How far these deviations may
extend without diminishing the ecphoric action or entirely
preventing it is a question to which a general answer
cannot be given. Each case must be considered in its
specific action.

We will now consider those influences, the ecphoric
action of which can in a certain sense only be described
as intermediate, inasmuch as they do not act on the
engram (A) itself which is " focused," but in some
engram (B) associated with the former. We find that
the ecphory of B, that is, the mnemic excitation /3, acts
ecphorically on engram A.

Already, in the section on the engraphic action of original
stimuli, attention has been drawn to the peculiarity of
engrams generated in an organism either simultaneously
or in quick succession. These are described as associated.
This association or connection is latent in the engram,
and only becomes manifest when owing to a strongly
marked association the ecphory of one engram unfail-
ingly affects the ecphory of the other ; the association
having been established by the frequent repetitions of
the simultaneous or successive engraphic stimulation;

To demonstrate the presence of such jnter media te
ecphory we ought, of course, to be able to produce such
associated engrams experimentally. But we have already
pointed out the difficulties of experiment on organisms
where no differentiation in the irritable substance as
such has yet taken place. If it is then difficult to create
by individual stimulation a single clearly defined engram,
how much more difficult will it be to produce two different
engrams simultaneously or successively associated ? The
reader is referred to the remarks on page 28. We shall,
however, meet with many cases of simultaneously or
successively associated engrams of lower organisms in
the later chapters of this book, especially in that part
which deals with the action of mnemic processes in onto-
genesis. For our present purpose we prefer to select
examples from the realm of the higher animals where,

ENGRAPHIC ACTION ON THE INDIVIDUAL 47

on account of the high specialisation of their irritable
substance in the shape of a nervous system, the general
production of engrams, and consequently also the simul-
taneous and successive production of different engrams,
is the readier.

It has already been described how in a young dog the
single simultaneous experience of stimuli — such as (i) the
optic stimulus of a man picking up stones and {2) the
tactual stimulus caused by stones which strike the dog
— suffices to generate for life two engrams, so closely
associated as to allow of the ecphory of the one engram
by the single repetition of the original stimulus belonging
to the other.

Any other dog without a hke experience reacts as the
first one did before the generation of the associated engrams,
that is, either it is indifferent at the sight of a stooping
man, or if it has been previously engraphically influenced
in the way of playing at recovering stones — the dog reacts
to the optic stimulus by the tension of muscles ready
for the jump, and by an accurate focussing of the man's
hand with the eyes in order to be able to judge the fall
of the stone immediately it is flung.

For examples of associated ecphory in vertebrates not so
highly organised as the dog we may consider the birds,
although they also are comparatively highly organised
creatures.

Lloyd Morgan has observed that the offspring of chickens,
pheasants, guinea fowls, water fowls, etc., on their emerg-
ence from the shell,and in the absence of parental guidance,
peck indiscriminately at things of a certain size, at grain,
small stones, breadcrumbs, wax matches, bits of paper,
beads, cigarette-ash, cigar-ends, their own toes, and those
of their compemions, maggots, bits of thread, specks on
the floor, their neighbours' eyes — anything and everything
about the size of a pea or less. But soon the optic engram
of an object becomes associated with an engram of taste,
and the bird ceases to peck at things non-edible or un-
pleasant to the taste. When Lloyd Morgan threw before
the chickens certain distasteful caterpillars conspicuous

48 THE MNBME

by alternate rings of black and golden-yellow, the larvae
of the moth Euckelia jacohicB, the young birds picked
them up at once, but dropped them immediately after-
wards. The one experience sufficed, in most cases, to
produce an optic and a chemical engram in an association
which became manifest on repetition of the experiment ;
for the optic stimulus induced the ecphory of the optic
engram, and indirectly that of the chemical engram, to
prevent the picking up of the striped caterpillars, while
other caterpillars, coloured brown and green, were taken
and readily devoured. After the second or third repetition,
both engrams were, in all cases, so well fixed and associated
that the black and gold striped caterpillars were either
no longer noticed, or the sight of them stimulated flight
and warning notes of disapprobation.

To produce clearly manifest and well-established engrams
in fishes, frequent repetitions of stimuli are necessary.
But Edinger's researches in that realm show most clearly
that such repetition of stimuli succeeds in producing
an association of engrams, and that with such simul-
taneously or successively produced engrams, the ecphory
of one effects the ecphory of the other. The engram
produced in the fish by repetition of th^ optic stimulus
of a man casting food upon the water becomes associated
with the optic and chemical engrams established by the
sight and taste of that food. The ecphory of the appearance
of the man suf&ces to act ecphorically on the associated
engrams, and liberates the reactions appertaining to the
latter, such as swimming to the side of the pond,
etc., even when the actual distribution of the food is
suspended. It could also be proved here that these
engrams survived a period of latency of four months
without disappearing or losing their associated or ecphoric
capacity. I have myself observed in the peculiar sucking
fish (Echeneis) in the Torres Straits, that such an engram-
association can be produced by a single experience, and
remain fixed for at least some hours or even for days.
It was sufficient to catch with a hook one fish out of the
large shoal of fishes unsuspiciously swallowing the food

ENGRAPHIC ACTION ON THE INDIVIDUAL 49

thrown to them for disgust of this feeding to be created
in the rest of the shoal for days. The experiment was
repeated many times, and always with the same result.
Edinger mentions similar observations on Abramis brama
and Idus melanotus, and further reports a number of
well-authenticated observations which prove that several
associated engrams remaining effective for days and weeks
can be generated in fishes by a single stimulation. Among
the most highly developed non-vertebrates, chiefly certain
insects such as wasps, bees, and ants, it is relatively easy,
as compared with experiments on thfi lower vertebrates,
to produce by a single stimulation engrams whose existence
may be determined with all necessary accuracy. Some
cuttlefishes respond to a like experiment in a similar
way. But the stimulus effect on these forms is always
restricted to specific stimuli, differing according to groups ;
and the lower we descend in the animal kingdom the more
frequent must be the repetition of the stimuli in order to
generate clearly demonstrable engrams.

We have entitled this portion of the chapter " ecphoric
influences," and not ecphoric stimuli. The title of stimuli,
however, would cover all the ecphoric actions enumerated
so far. But now we pass on to consider influences which
cannot exactly be described as stimuli, although their
character is an undoubtedly ecphoric one.

We may take our first example from a very common
experience. Suppose I am in the habit of having my
first meal at eight o'clock in the morning, my second at
noon, and my third at eight o'clock in the evening, the
complex stimuli connected with each partaking of food
produce the following reaction amongst several others.
The sight and taste of the dishes are accompanied by a
peculiar reaction of our sensory areas which we describe
as hunger or appetite, and which under ordinary circum-
stances is absent in healthy well-nourished individuals
during the intervals between meals. If I now suddenly
interpose between these meals two slight repasts — one
at eleven and the other at five o'clock — I shall experience
at first some disinclination. But I force myself to them,

4

50 THE MNEME

let us say, on the doctor's recommendation, and adhere
to the order for half a year. If at the end of that time
I try to dispense with the eleven and five o'clock meals,
a strong feeling of hunger makes itself distinctly felt
at those times. Apparently, the lapse of a definite period
acts ecphorically on this reaction of my sensory areas.

" X™6 " a-lso apparently influences other reactions
of our body. We need not dwell on respiration and
pulsation, because the intervals between the single reactions
are so short that it appears unlikely that a return of
the irritable substance into its state of indifference can
actually take place, and that these phenomena can be
regarded as mnemic at all. The real nature of the rhythm
may be explained equally well in other ways, although
the manner of its expression in various species of animals
is probably influenced by inherited engrams. I shall not
however, discuss this question further in our analysis.

On the other hand, the time interval connected with
the periodical maturation of the ovum and the changes
in the mucous membrane of the uterus, which we call
menstruation, bears distinctly the character of an ecphory.
In all periodical phenomena in the animal and vegetable
kingdom, whether acquired or inherited, time is commonly
supposed to determine and regulate the appearance
and disappearance of reactions. This may be explained
by the help of a well-known example which, however,
happens to be a case of inherited engrams ; but as for
this present analysis the nature only of the ecphoric
influence matters, and not the origin of the engrams, the
choice may be permitted. Most plants of the temperate
and frigid zones possess an annual period, that is, a
definitely determined alternation between the pause in
growth and the progress of vegetation. The turning-
points are, as is well known, very distinctly marked in
our deciduous trees by the fall of leaves in autumn and
the sprouting of new buds and leaves in spring. This order
is unmistakably connected with the cUmatic periods
we call seasons, and naturally depends primarily on the
geographical latitude, and among other things on the

ENGRAPHIC ACTION ON THE INDIVIDUAL 51

height above sea-level, vicinity of mountains or sea,
predominating air currents, and peculiarity of position.
We need not labour the point that in the frigid and
temperate zones the change of temperature is the determin-
ing factor in the dependence of the vegetation periods
on the seasons. In the tropics, however, the relations
between the seasons and the vegetation periods depend
on variation in humidity. Examining the plants of
our own zone, we find that in the case of a few the change
from cold to warm determines in each instance the
transition from rest to new growth. Some of our shrubs
will certainly sprout in winter, if in January or February
we have mild weather with plenty of sunshine. Then
snowdrop, crocus, scilla, auricula, and daphne will begin
to flower, the honeysuckle will unfold its leaves, and the
Ulac-trees show the first shimmer of green. All these
are plants which can be made to grow and blossom
prematurely in a forcing house. An attentive observer,
though he may be but a layman in natural science, will
soon notice that the various plants behave differently
under forcing. Whilst a great number of the most varied
plants can be regularly forced without difficulty — ^mostly
those which sprout early in spring — others offer more
resistance to the process, while on some plants forcing
has but a feeble effect. Observations carried on for
years on plants grown in the open air gave similar results.
The Munich winter of 1899-1900 was very long and very
cold. Heavy snowfalls and severe frosts occurred in
March and in the beginning of April. One night in early
April the thermometer fell to 15° C. Not until after the
middle of April did the warmer weather set in. Con-
sequently, vegetation in the Isar valley near Munich
was still very backward in April, Snowflakes {Leucojum
vernum) and crocuses did not bloom in our shaded garden
until mid- April had passed. In donning their vernal robes
most bushes were late ; Lonicera tartarica was clothed
about mid- April, and Spanish lilac only towards the end of
the same month. The leaf buds on most branches of a
certain copper beech opened out on the first day of May,

52

THE MNEME

The winter and spring of 1901-1902 offered a strong
contrast to the corresponding seasons of 1899-igoo.
The winter was mild throughout, the spring without
severe frost or snowfalls. The temperature was fairly
even and moderate, and now and then we had spells of
sunshine. In this year the first snowflakes bloomed in our
garden on the 17th of March, the first crocuses on the
20th of March, that is, nearly four weeks earlier than
in the year 1900, The first green leaf of the Lonicera
bushes appeared on the 20th of March, and the lilacs started
to unfold their leaves on the loth of April. The before-
mentioned beech began to leaf this year on the 23rd of
April, that is only one week earlier than in the year 1900,
a fact worthy of note in a year when the majority of the
other plants growing under the same conditions were
three or four weeks ahead of the corresponding phase
of vegetation in 1900.

Observations continued for several years in order to
illustrate a fact which had long been known to botanists
and gardeners gave convincing results. The following
table shows that the unfolding of the leaves of our beech-
tree between the years 1900 and 1905 varies as to time
within the narrow limits of the 22nd of April and the
4th of May ; thus evidencing a far higher degree of inde-
pendence of the annually changing climate influences than
most other plants: —

Year.

Leucojum vernum
to blossom.

The Crocus
commenced
to blossom.

The Leaves
of Lonicera

tatariea
commenced

to unfold.

The Leaves

of Syringa

vulgaris

commenced

to unfold.

The Leaves

of Isolated

Specimens of

Fagta silvaiica

commenced

to unfold.

1900

15 April

17 April

17 April

17 April

I May

igoi

—

—

—

—

—

1902

17 March

20 March

20 March

10 April

23 April

X903

25 February

20 March

8 March

26 March

4 May

1904

13 March

19 March

25 March

8 April

22 April

1905

9 March

II March

—

17 April

23 April

ENGRAPHIC ACTION ON THE INDIVIDUAL 53

It is clear that the periodicity of a number of plants
does not stand in immediate and pre-eminent relation
to the lack or excess of vernal warmth, but that a second
factor — the element of time — has a predominating influ-
ence. Even in those plants which can be forced the time
factor makes itself known. In very few plants, indeed,
is it possible without considerable difficulty to overcome
the influence of the time factor. Plant-physiological
researches and experiments by nurserymen show that
even those plants which are relatively easily forced resist
in the first part of their resting period the stimulus of
temperature, even, of all external influences, and only
reach a state in which forcing becomes possible after
the lapse of a definite period which differs according to
species and geographical distribution.

In order to observe the direct simple action of the time
factor, one must eliminate as far as is possible the stimulus
of temperature. This was done by potting in the spring
of 1903 a one- year old beech and cultivating it in an even
temperature side by side with two beech seedlings, likewise
potted. From the ist of September onwards the plants
were kept in the room in order that they might be protected
from the night-chills. Nevertheless, the leaves commenced
falling on the 22nd of September, and were all off by the
15th of November. All through the winter these three
branches stood in a room of the same temperature both
day and night, and were watered with water of moderate
temperature. No unfolding of leaves showed itself of
any of the trees until the ist of May, when the beech, now
two years old, started. Of the two seedlings, now one
year old, one began to unfold its leaves on the 25th of
May, but the other started only in mid- June. The
retarded, leafing might be explained by the harm the
plants suffered through having been entirely withheld
from the winter cooling. It is known that this interfer-
ence can be endured without harm by only a few plants
of the temperate zone. In this connection it is interesting
to note how the hereditary character of these dispositions
can be deduced from the retention of the periodicity of

54 THE MNEME

those two seedlings, which during their individual lives
never were subjected to the least periodical influence.

But what do we mean by the expression " time factor " ?
Or, as we are not desirous of discussing the theory of
knowledge as such, how can we form a conception of a
time period which in its influence will resemble or represent
the action of a stimulus ?

All life processes as experienced appear to us to be
regulated by time. We find that the state of excitement
of any specific organism requires definite time to propagate
itself through equal extents of irritable substance. The
beginning and course of all other reactions is. under given
circumstances, equally determinable in time. Is there not
here some clue to the nature of the time influence ? To
a plant or an animal a time period means the elabora-
tion of a definite number of life processes in its system.
A man, who knows the average rate of his pulse and
respiration, could without reference to a chronometer
fairly correctly estimate the passing of minutes and even
of hours, if he were willing to go to the trouble of the
prolonged counting.

A prisoner in a subterranean prison, artificially lighted
and heated, denied communication with the outer world,
and receiving his food at quite irregular intervals — ^by
CEdculating the rate of growth of his nails and hair, by
reference to pulse beats, thus making his own body serve
as a watch — could register the run of the weeks and months
and years with approximate accuracy.

The chronometer of the organism, which one might
call its " body watch," is thus regulated by the rate of
speed of its life processes. But how does the organism
read the close of a time period from this chronometer
without conscious counting ? Or, to speak less in simile,
how is it that a quite definite reaction sets in after the
close of a specific series of life processes ? Simply because
at the end of a definite series of metabolic or other life
processes, a state of the organism results which partially
or altogether corresponds to the state which ruled at the
time of the production of a definite engram, and by the

ENGRAPHIC ACTION ON THE INDIVIDUAL 55

repetition of which state this engram becomes now
ecphorised. A Central European beech, which stands
in full vegetation from May until September, reaches in
the latter month an organic state which acts ecphorically
on that engram whose successive reactions consist of
food circulation from leaves into branches and root-stems,
and of the fall of leaves. This ecphory takes place in the
beech in autumn irrespective of those specific influences
of temperature which rightly are expected to act at that
time of the year, and which by abnormal mildness fail
to come into play. We must remember, however, that
otherwise normal temperature would also act ecphorically
on the engram.

We conclude that the time period as such does not act
ecphorically, but that the ecphory is due to the appear-
ance of a definite state associated with the respective
engram, and that this appearance is determined in time,
in so far as it takes place on the conclusion of a definite
mmiber of life processes which may be estimated from
the moment chosen as the starting-point.

That these processes are the result of the ecphory
of an engram is proved by the fact that by the frequent
repetition of a new and appropriate stimulus leaf-fall
can be associated with some earUer or later engram of
metabolism. I term such engrams and their ecphory
" chronogeneous." From some cause that is still unknown
to us, this can only be effected with great difiiculty in
the case of the beech, but in the case of most other plants
it is sufficient to expose them to a different temperature
and light for a number of years for the engram of, say,
leaf-fall, or the engram of forcing to associate itself with
some other chronogeneous engram. Numerous instances
are furnished by the science of plant-acclimatisation.

The ecphoric influence which the appearance of a
definite phase of development in the life history of an
organism exercises on a definite engram has to be classed
with this chronogeneous ecphory. Here, also, it is a
case of the association of a certain engram with a specific
phase of organic development. Further, this ecphory is

56 THE MNEME

based on the association of various engrams ; but that
conception we shall analyse more minutely in the second
part. We may simply note that the time factor here
is pushed more into the background than in the previous
cases of chronogeneous ecphory. We shall describe such
a state as " phasogeneous ecphory," which means that
on reaching a certain phase of development a state of
the irritable substance has been generated which acts
ecphorically on a certain engram.

Such expressions as " chronogeneous " or " phaso-
geneous " ecphory will be simply used for the sake of
ready reference, and not with any idea of making them
specific categories distinct from other ecphories. It is
characteristic of them, as well as of all other ecphories,
that the partial return of an energetic condition acts
ecphorically on the engram-complex of the total situation.
But these questions will receive more detailed treatment
in the second part of this book.

CHAPTER III

ENGRAPHIO ACTION OF STIMULI ON PROGENY

In our analysis of mnemic phenomena we have preferred
to consider engrams acquired during the individual life
of the organism under examination, and only occasioneilly
have we considered engrams inherited by the organism
from its ancestors. To ignore these entirely was impossible,
for they exist in each organism from the ovum stage
onwards, and interfere in manifold ways with the course
of all our experiments.

Proceeding now to their closer analysis, we have first
to consider a point of fundamental importance, namely,
whether engrams persist through the individuality phase
in which they have been generated into the immediately
succeeding phases.

To answer this question we will consider various crucial
tests taken from four distinct groups of observations
which, to our mind, demonstrate clearly the inheritance
of engraphic actions.

In the first group the inherited engram, or rather, the
inherited engram-complex, manifests itself by reactions,
which concern primarily, though not exclusively, the
sphere of instinct. The experiments cited here are those
which P. Kammerer elaborated.

The so-called fire newt, Salamandra maculosa, which
is viviparous, ordinarily gives birth to larvae measuring
on the average 25 mm. and carrpng gills. The broods
may number anything from fourteen to seventy-two.
These larvse are deposited in the water, and there pass a

long transitional stage, until after several months they

57

58 THE MNEME

lose their gills, leave the water, and completely change
into land newts. By change of* conditions Kammerer
succeeded in forcing the females to retain their offspring
for a longer time in the uterus. Then, by repetition of
the coercion, these retarded parturitions became habitual.
In order to induce the females, so far bearing normally,
to retain their offspring in the uterus beyond the usual
time, Kammerer kept them from the water-trough in
which ordinarily they would have deposited their larvae.
The keeping of the animals in a cool temperature, which
rarely was higher than 12° C, proved of assistance.

By artificial conditions Salamandra maculosa can thus
be compelled to retain her offspring in the uterus until
they are fully developed, as is normally the case with the
black Alpine newt, Salamandra atra. This remarkable
change is effected in four stages : — (i) Numerous larvae,
25-30 mm. in length, are deposited on land instead of
in water. (2) Later, a smaller number, but still further
advanced in development, are similarly deposited. (3) A
still smaller number of larvae, seven at the most,
with reduced gills, either ready for their immediate
metamorphosis into the adult newt, or just past it, are
deposited. (4) The number of individuals in a brood
grows gradually less from one gravid period to another,
until, as in the case of the Salamandra atra, the number
of offspring remains constant at two, that is, one foetus
in each uterus.

Kammerer now reared a number of young ones, born
after their mothers had reached the highest stage of
habitual late parturition, and mated them amongst
each other. He then subjected the fertilized females
of this second generation to normal conditions during
their pregnancy by supplying them with a water-trough
and sufficient moist air, and by maintaining them at a
moderately high temperature. Although there was no
longer any external coercion to a late parturition, the
creatures departed from the normal breeding mode of
their species, and without exception from the beginning
made late and sparse births. It is true that they did

ENGRAPHIC ACTION ON PROGENY 59

not bear fully metamorphosed newts, but their newt
larvae stood in all cases nearer to their metamorphosis
than the normal larvas. Further, the number of each
brood had decreased from the average thirty or forty
to five, four, or two larvae.

If, on the basis of previously enunciated views, we
analyse this effect of change in external conditions in-
herited from one generation to another, we find the
following elements : —

I. A primary state of indifference. By this we refer
simply to the state of the organism at the commencement
of our observations and experiments. In the case before
us it means the state of the maternal generation before
coercion.

II. Stimulus to be tested for its engraphic action. This
is of a complex nature, and may be defined compre-
hensively as the coercion of the females to retain the
young in the uterus beyond the normal time of pregnancy.
This end was effected partly by keeping the animal in a
relatively dry, cool state, but mainly by the withdrawal
of the water-trough. When at each subsequent pregnancy
this coercion was repeated, the effect of its action increased
from case to case until the retarded parturition finally
reached the fourth stage. We conclude from this that
the complex coercion-stimulus effected an unmistakable
engraphic action on the individuals of the first generation.
So far our observations warrant us in speaking only of
engraphic action in the case of individuals directly subjected
to the stimulus, and with these, of course, it is a case
of an individually acquired and not an inherited engram.

To establish proof of the inheritance of the engram
we must turn from this generation on which the stimulus
has directly acted as original stimulus to the next genera-
tion, and keep the animals under conditions where they
are withdrawn from those influences which have acted
engraphically on the mother.

III. A secondary state of indifference. This is the state
in which the second generation, born of the habitually
late-bearing mothers, remain from their first pregnancy

6o THE MNBME

until that moment which corresponds to the normal
bringing forth of the young of Salamandra maculosa.
During this time they do not differ in anything from
newts born of normally-bearing mothers.

IV. Ecphorid influence. Proof that the repeatedly
practised forcing of the mothers to retain their young
ones in the uterus beyond the ordinary time has not
only acted engraphically on them, but also on their ofE-
spring, is furnished by the following data: — When the
gravid offspring reach the phase of the normal parturition,
the delivery is retarded, and late parturition regularly
takes place instead, although the mothers of the second
generation during their pregnancy were kept under
normal conditions of moisture and temperature, and with
a water-trough at their disposal.

The external means used to cause the first generation
to retain their young in the uterus were not applied to
the second generation. The engram, ecphorised on reaching
the corresponding stage of pregnancy or, as we may
say, the mnemic excitation ecphorised phasogeneously,
manifested itself in this case by the retardation of the
normal act of parturition.

On the other hand, in the case of the Alpine newt,
Salamandra atra, the inherited engram manifests itself
during its phasogeneous ecphory by a positive reaction,
and not by retardation. Salamandra atra normally
deposits fully developed land newts on dry soil. By
keeping the animal in a relatively high temperature
and by supplying it with abundance of water — means
very different from those used with the spotted newt —
Kammerer succeeded in introducing premature parturition
in the Alpine newt, gradually increasing the brood and
finally making the change habitual. The induced change
was transmitted to the offspring which, kept under con-
ditions normal to their species, without exception produced
a more numerous but less developed progeny at a time
earlier than the normal act of parturition. In this case,
the phasogeneously-ecphorised engram manifested itself
not by the retardation of a process normally due, but

ENGRAPHIC ACTION ON PROGENY 6i

by its acceleration, that is, by a relatively premature
parturition.

In a similar way the ecphory of the inherited engram
expresses itself positively in the case of the change of the
development and instinct in the Axolotl described by
Miss V. Chauvin, and to which reference will be made
later, as well as in the manifold changes of instinct in
Butterfly-caterpillars and Beetles, Lymantria, Gracilaria,
Phratora, experimentally brought about by Pictet and
by Schroder. In a later passage we shall treat in detail
the successful experiments in heredity made by Kammerer
on the Obstetric Toad, Alytes obstetricans.

For the moment, however, leaving those cases of inherit-
ance which manifest themselves in their ecphory principally
by reactions of instinct, as also by certain morphological
and physiological characters in the newly born, let us
concern ourselves with equally convincing researches in
which the inherited engram manifests itself by plastic
reactions in the bodily development of the offspring.
Let us begin with the experiments made by E. Fischer
on the Tiger Moth, Arctia caja, experiments which leave
nothing to be desired as regards the unequivocal nature
and force of their evidence. Earlier than Fischer, Standfuss
had made similar experiments on butterflies, with similar
results.

Later, Schroder, Pictet and Tower, among others,
obtained analogous results from butterflies and beetles.
Stimuli of different kinds were applied in these experiments.
In those of Fischer, stimuli of temperature were applied,
the pupae being kept at the low temperature of 8° C. The
reaction produced by this stimulus manifested itself as a
plastic one, by alterations in the pigmentation of the wings
of the emerging imago.

I. The primary state of indifference was traced with
sufficient accuracy by Fischer in this experiment by
keeping permanently at a normal temperature half of
the number of pupae collected for these trials.

These pupae, with the exception of five from which no
images emerged, gave moths that showed no appreciable

63 THE MNEME

alteration in colouring and marking. Neither the brown
spots of the frontal wings nor the black ones of the back
wings showed any deviation from the normal. The
development of such normally coloured moths may in
this case rightly be described as the primary state of
indifference.

II. The stimulus, whose engraphic action we are now
examining, consisted in Fischer's experiments of an inter-
mittent decrease of temperature until it reached 8° C.
This stimulus was apphed to the brood in its pupal
stage, with the result that the emerging moths " showed
aberrant colouring of the wings. Seven of the total
number of forty-eight pupae died. The aberration lay in
an enlargement of the brown spots on the front wings
and the black spots on the back wings. In some of the
male specimens the dark colouring affected the whole
of the wing." — " The moths also showed a correspond-
ing change on the lower part of their bodies." An
unmistakable melanotic aberration had taken place.

III. Secondary state of indifference. This we defined as
that state where the excitation of the engraphic substance
produced by the stimulus had subsided, having effected
a latent engraphic change. Exactly when this state
supervened in the generation of Arctia caja, exposed
to the stimulus of cold, need not here specially concern
us. The point is that this secondary state of indifference
persisted until, on the application of the stimulus, the
reaction was given. During the time of development
from the egg to the pupal stage, no reaction deviating
from the normal can be observed, either in growth or
in any other biological phenomenon.

Of the descendants, we need but to consider the offspring
of two moths which had been affected melanotically by
the stimulus of cold. The male had been much altered ;
the female less so. Fischer gives a drawing of them
both. From the pairing of these two abnormal speci-
mens resulted 173 pupae, which were kept at a temperature
of from 18° C to 25' C.

VI. Ecphory of the engram. In the absence of that

ENGRAPHIC ACTION ON PROGENY 63

stimulus of cold which had acted engraphically on the
parent generation, the case proved to be one of phaso-
geneous ecphory. For when the pupation was finished
seventeen out of the 173 specimens showed melanotic aber-
ration entirely along the Une of aberration in the parents.
Comparison with the parents shows that " the offspring
represent generally a combination of parental character-
istics. The marking of the male preponderates in some,
that of the female in others. It is worth emphasising,
however, that the aberration had taken place most
strongly in the male moth. Further, in these descendants
the under side of the wings is altered similarly to the
upper side."

The stimulus of cold on the parents not only generated
engrams which manifested themselves in descendants by
morphogenetic reactions, but generated also others, which
found expression in other reactions. For the seventeen
affected specimens were amongst the latest of the emerging
moths, whilst those which came out at the beginning were
entirely normal. In the first edition of this book it was
suggested that the stimulus of cold might generate another
inheritable engram, namely, retardation of the rate of
development. This idea has since been confirmed, for
Pictet discovered that accelerated' pupation regularly
accompanies melanotic aberration, whilst retarded pupa-
tion accompanies albinotic aberration.

Here then are two perfectly clear cases of engraphic
effects, which are open at any time to be tested anew.
They are representative of a great series of related cases.
Let us now turn to a third group of cases demonstrating
inherited engraphic changes in plants.

It has long been known that some trees, when trans-
planted from the temperate zone into moist tropical
regions, gradually relinquish their periodical faU of leaves,
and change from deciduous into ever-green plants.
Bordage, during his twelve years' stay at Reunion,
investigated the question whether this alteration climatic-
ally induced might not be transmissible by seed. The
idea of grafting was not entertained. He found a suitable

64 THE MNEME

object for his experiments in the peach-tree. Plants
grown in Reunion from European peach kernels continued
for ten years to drop their leaves annually at the ordinary
time, even when cultivated in the hot coast districts.
At first they remained entirely leafless for about a month
and a half, but gradually the period was shortened.
After ten years some specimens were so much changed
that they continued in leaf the whole year through.
But it was not until after twenty years that they reached
a state in which they might be described as practically
evergreen plants. Bordage speaks of this as " subpersist-
ance du feuillage." If now the seeds of these evergreen
trees are sown, the resultant plants are evergreen like
their parents. This occurs even when the seeds are sown,
not in the hot coast lowlands, but a thousand metres
above sea-level, where those peach-trees, the descendants of
parents not changed by the climate, permanently retain
a periodical fall of leaves. A more detailed enumeration
of cases is given in the Problem of the Inheritance of
Acquired Characters.

We now arrive at the conclusion that the numerous
stimuli which continually act upon an organism have not
only synchronous and acoluthic, but very frequently
also engraphic effects. Further, it has been shown that
these engraphic effects extend beyond the individuality
phase into later phases of the continuous line of develop-
ment, that is, they are inherited. From this it follows
that each organism, linked up as it is with the history
of miUions of years and consequently representing the
millionth or billionth phase in a continuous line of develop-
ment, must contain a great many such engrams ancestrally
acquired which have been transmitted to it through the
generations. In our investigation of organisms, do we
find in their irritable substance properties that possess
the character of inherited engrams ? Whether they are
inherited or not will, of course, be determined by the
observation of several generations. But it will be exceed-
ingly difficult to decide whether we are dealing with an
engrani, that is, a latent residue of a previous stimulation.

ENGRAPHIC ACTION ON PROGENY 65

The inherited engrain is the product of a stimulation
upon a previous generation. We therefore have to deal
with an historical occurrence, and as a rule in such a
case the experimental method is denied us. Even if
in thousands of organisms we show that we are able to
generate engrams which become inherited, this by no
means proves that the inherited dispositions we find in
organisms are actual engrams.

We have arrived at a critical point in our analysis, for
we are about to enquire into the nature of the dispositions
hereditarily transmitted from generation to generation.
Hitherto, we have had the solid ground of immediate
experimental experience beneath our feet. For a time
our further journey will be in the realms of conjecture.
Before we enter fully into the matter, perhaps we may
be permitted a short but somewhat necessary digression.

To begin with, every conclusion which tells us something
really new, which adds something to our knowledge, is
after all an inference from analogy. The mathematical
and purely logical conclusions termed " necessary " are
only transformations of what is already contained in the
premises. They do not tell us anything new in fact.
Experience alone gives us that which is new. And the
method is by analogical inference. Hence originate our
conceptions of time and space, the mathematical theorems
and the fundamental propositions of science. That a
stone thrown up in the air will fall to the earth we
know only per analogiam. For knowledge of the existence
of the force of gravity is not innate in the human mind,
but is an inference drawn from a great mass of analogous
experiences. The same may be said of the law of the
conservation of energy, and of every conception of natural
science.

Now we can divide the things which present themselves
to the mind in causal series into two groups — those which
repeat themselves or are capable of being repeated, and
those which appear but once and cannot be repeated.
Strictly speaking, only those occurrences which can be
repeated are matter for experiment. The laws of falling

5

66 THE MNEME

bodies find continual confirmation in new experience.
The presuppositions and conditions of the test experiments
may be varied indefinitely, but the same result always
follows. Conclusions which can thus be confirmed continu-
ally by immediate experience acquire for us the character
not merely of probability, but of inevitability.

Occurrences which cannot be repeated escape the direct
control of fresh immediate experience. It cannot be
proved by new direct experience that the Triassic cretaceous
rocks have been formed by the precipitation of solid
ingredients from a liquid medium. An experimental
imitation of the process is merely an imitation, not a
test experiment. But the fact that in this deposit we
find creatures like the Echinoderms known to be living
at the present time exclusively in the sea, whereas
terrestrial creatures are absent, considerably increases
the force of the probability that the chalk strata were
originally laid in the sea. But the sceptic, clamouring
for direct experimental evidence, is not easily satisfied
even by the most cogent indirect reasoning. In this
case, he may perhaps argue that the fact that present-
day Echinoderms are exclusively marine animals does
not prove anything in respect of the Echinoderms of the
Trias. But that non-recurring physical and biological
phenomena may nevertheless warrant conclusions con-
vincing to any normal human being may be demonstrated
by any number of examples. Is there any thinking man
who seriously questions the idea that the fossilised animals
and plants have once actually lived ? Or that the fossilised
vertebrates possessed nerves ? Is he justified in rejecting
the idea because the fact cannot be proved by direct
experimental evidence ? It has recently become the
fashion among a certain group of biologists to deprecate
the value of the historical method because of the indirect
conclusions involved. It would be interesting to know
how far these men of science would allow their scepticism
to affect their reading of social history. Is the account
of Erasmus and the Reformation to be rejected because
it is not possible to set out an experimental repetition ?

ENGRAPHIC ACTION ON PROGENY 67

Why should natural science be denied the valid use of the
historical method ? It is a comfort to know that the
majority of men, learned and unlearned, will continue
to devote thought and attention to mere historical, non-
recurring phenomena.

The problem, whether vital and physical phenomena
can be made to recur at all in identical fashion need
not here be investigated. It is admitted that many
occurrences in the inorganic world lend themselves to
such uniform repetition that, as far as the resvilts are
concerned, the deviations are hardly appreciable, and
need not be reckoned. But in the organic world the
deviations in each occurrence of an event are much greater.
No single organism can serve in repeated experiments
in exactly the same way. And in the case of different
organisms, for instance, of two individuals of the same
species, the innate differences may be considerable.
Nevertheless, we may speak of recurring phenomena also
in the organic world, as long as we limit the expression to
those characteristic features, in comparison with which
the infinitesimally small deviations at each recurrence
do not count. It is, however, well to remember that
in the strict sense the identical recurrence of organic
events is impossible. We fully acknowledge the superiority
of the direct experimental conclusion over the indirect
historical conclusion, but we shall not give up the latter,
since it affords us the sole key to the understanding
of non-recurring historical phenomena, and because its
results may, in favourable cases, possess such convincing
force of evidence as to approach the results obtained by
direct experimental methods so nearly that the difference
of value becomes infinitesimally small.

By the historical method, which of course involves
logical process, we conclude that fossils are the remains
of animals and plants that formerly lived, and that they
are not " freaks of nature." This is a demonstrable
conclusion. The perfectly accurate assertion that the
laws of mechanics are, by the experimental method,
still more definitely demonstrable does not a£fect our

68 THE MNEME

conclusion. The point is that conclusions by either method
are valid. Appl5dng these considerations to the problem
of the inheritance of engrams, it is clear that as regards
the genesis of these engrams we have in the majority
of instances to deal with non-recurring events. This
is indeed always the case when the engrams have
been acquired by ancestral generations which differed
considerably in form and function from the animal and
vegetable generations now living. All these age-long
acquisitions are beyond the direct experimental method
of proof. They can only be established by an inferential
proof.

If now we set forth our argument — that the great
majority of the dispositions inherited by organisms are
to be regarded as engrams, asking for its acceptance on
the grounds of strong probability — we shall have first
to enquire into those particular features by which an
engram is recognised.

The surest signs of an engram are derived from the
observation of the phases of its genesis. We note (i) the
state of the organism before the existence of the engram,
that is, the primary state of indifference ; (2) the action
of the engraphic stimulus ; (3) the secondary state of
indifference, that is, the phase of latency; and (4) the
ecphory or the phase of manifestation.

In the historical engrams which concern us now, the
action of the engraphic stimulus belongs to the past,
The organism as presented to us for investigation is
already in its secondary state of indifference. Our
examination, therefore, can only deal with the phases of
latency and manifestation. The argument that in this
case we are really concerned with an actual engram must
proceed on two considerations. First : On the fact that
it concerns properties of the organic substance which
at times are latent, and at other times are manifest ;
and secondly : on the manner in which the transition
from the phase of latency to the phase of manifestation
is made, that is, on the evidence thf^t tbis transition b§ar§
the character of an ecphory,

ENGRAPHIC ACTION ON PROGENY 69

Now, all the inherited properties which concern us
here are in a state of latency, out of which at each recur-
rence of the hberating influence they pass into the corre-
sponding state of excitement.

The reaction by which an engram becomes manifest
to us naturally does not differ as such from any of the
reactions effected by an original stimulus. The difference
lies rather in that which evokes the reaction. But can
we learn from the liberating influence itself whether
its action is ecphoric, or whether it is indeed an original
stimulus ? Here we venture on grounds of probability.
As we have already seen, the existence of an engram is
recognised by the fact that the unaltered original stimulus
is no longer requisite for the production of the correspond-
ing reaction. Instead, there may be only the action either
of the quantitatively or qualitatively changed original
stimulus, or of a stimulus acting ecphorically on an
associated engram, or the mere lapse of a definite time
period, that is, chronogeneous ecphory, or, finally, the
initiation of a definite development phase in the continuous
line of the successive generations, that is, phasogeneous
ecphory.

Now, in all organisms, protozoa, plants, and animals, we
meet with an extraordinarily large number of excitation-
dispositions, the corresponding stimuli of which have in
all probability to be classed in one or other of the above
mentioned categories.

The first named categories of ecphoric influence, where
the merely quantitative or qualitative change of the
original stimulus comes into play, are no less important
and imcommon than the others mentioned. But as
they more or less resemble the original stimuli, it is clear
that for our present exposition they are of far less value
than are the cases of associated chronogeneous and
phasogeneous ecphory. Still, what they offer is by no
means unimportant. In many species of birds the reaction
of pecking at first sight at grains and other minute objects
is inherited. That this is a case of an inherited engram
on which the mere optic stimulus of the object acts

70 THE MNEME

ecphorically seems to me most probable. It sometimes
happens, however, that the optic stimulus alone is not
sufficient t6 produce the reaction, and with many a young
chicken and pheasant, bred in the incubator, it takes
some time for the optic stimulus of the scattered grains to
produce the reaction of pecking. This reaction can,
however, be initiated by the example of older and more
sophisticated chickens, or by touching the grains before
the eyes of the young creature with the finger-nail or
a pencil, in imitation of the pecking of the hen. This
is especially the case with young ostriches bred in the
incubator, which, according to Claypole, do not pick up
the food unless one first touches the ground where it
is scattered. Of the explanations of this phenomenon,
the most probable, in my view, is that it is the ecphory
of an inherited engram. The engram is that of which
pecking is the corresponding reaction ; the ecphoric
stimulus is the recurrence of a primary stimulus somewhat
altered qualitatively. So instead of the demonstration
of pecking by the mother hen, we have the touching of
the food with fingernail or pencil — a case of vicarious
ecphory.

Still more convincing as to the engraphic character of
many inherited dispositions, and the ecphoric character of
the influences liberating them, are the observations
made on young birds which on the first contact of their
beaks with water, were induced to go through the
whole ceremonial pretence of a bird-bath. Lloyd Morgan
reports in the fourth chapter of the book previously
cited several cases, one of which, observed by Charbonnier,
is here given. " A magpie about five weeks old, which
he had reared from quite an early stage of its life, when
placed in a cage and suppUed with a pan of water, made
one or two pecks at the surface, and then, outside the pan,
without entering the water at all, proceeded to go through
all the gestures of a bird bathing — ducking its head,
fluttering its wings and tail, squatting down, and spreading
itself out on the ground. It afterwards and by degrees
acquired the habit of real bathing, and seemed always

ENGRAPHIC ACTION ON PROGENY 71

anxious for a bath in rainy weather." The strangeness
of the bird's conduct becomes intelligible when regarded
in the light of engram-inheritance. The stimulus of the
contact with water, sUght though it was relative to body
area, acted ecphorically.

The cases so far given are instances of more or less
direct ecphory. We may meet, however, with numerous
inherited dispositions whose manifestation depends on
indirect ecphoric influences. The next case is one of
many which are intelligible only on the assumption of
several associated engrams. P. Huber gives an account
of a caterpillar which in a series of about a dozen processes
constructs for its pupation a very complicated cocoon. A
caterpillar that had reached the sixth stage in the spinning
of its cocoon, when placed in one not so far advanced,
immediately took up the work at the earlier stage and re-
peated the third, fourth and fifth processes in the making
of the cocoon. But when Huber placed a caterpillar
from a cocoon in its third stage of development into
one which had reached the ninth stage, the animal was
unable to take advantage of the work already done and
proceed from the ninth stage, but neglecting the inter-
mediate stages, had to start from the third at which it
had been interrupted, so that the fourth to the eighth
stages of the cocoon were spun twice. The case finds a
natural explanation in the idea that the complicated
act of spinning is the manifestation of a chain succes-
sively associated inherited engrams. The ecphory of each
engram acts in its turn ecphorically on the next successively
associated engram.

In the engrams hitherto discussed, the reactions mani-
fested themselves in muscular contractions such as pecking,
the stretching of wings of the young bird, and the compli-
cated spinning activity of the caterpillar. It has already
been emphasised, but I should like to lay stress on it
again, that it is absolutely immaterial to the fundamental
questions occupying us now whether the reactions under
observation by which the states of excitement of the
irritable organic substance become manifest to us consist

72 THE MNEME

in muscular contractions, or in changes of the cell tur-
gidity, or in processes of secretion and metabolism, or in
cell divisions and other processes of growth, or, finally,
in sensations whose existence can only be inferred by us.
We now turn to inherited dispositions whose correspond-
ing reactions are produced by time influences, that is, by
the lapse of a definite time period. These reactions have
already been classed under the heading of chronogeneous
ecphory (p. 55). The manifestation of inherited dis-
positions by chronogeneous ecphory may be discerned
in a very great number of animals and plants. Mention
need only be made of the periodic ovum maturation in
woman and of the mating periods of most animals.
The reactions by which these engrams become mani-
fest consist primarily in processes of growth. The
" migration impulse," innate in so many species of birds,
is the motor reaction of an inherited engram, a reaction
produced by chronogeneous ecphory.

The vegetation periods of the plants are phenomena
in which chronogeneous ecphory plays a very important
part. This is most obvious in plants which resist forcing,
that is, in those plants where the reaction of growth
admits of little interference. When a beech kept at an
even temperature in a heated room withers and drops
its leaves in November, despite the absence of that direct
stimulus of cold which at other times is necessary to this
reaction, it is clearly a case of a chronogeneous engram
manifesting itself in a chronogeneous ecphory. For
otherwise, in the absence of the stimulus of cold which
usually produces the leaf-fall, we should expect the beech
to retain its foliage. That this is a case of an inherited
engram is readily proved by using for these experiment
seedlings which have always been kept at an even tempera-
ture, and which, having been grown from seed, have
never yet been engraphically influenced in this particular
way during their individual lives.

The matter, however, is somewhat complicated in
so far as natural selection has determined the specific
annual periods of those plants, interference with whose

ENGRAPHIC ACTION ON PROGENY 73

growth by way of forcing is almost an impossibility.
For it must be of great importance to plants sensitive to
frost not to be tempted by the warmth of an early spring
to unfold their buds prematurely, and thereby risk the
blight of later frosts. The case of the annual period,
therefore, is by no means so clear as the one of the
diurnal period next to be discussed, in the hereditary
determination of which natural selection apparently has
little to do. The diurnal periodic leaf movements of plants
are ecphorised chronogeneously for some time after the
cessation of the light-stimulus that normally liberates
them. That this recurrence of the diurnal period, which
may be observed in plants kept in constant darkness or
in constant exposure to light, is a case of inherited, and not,
as Pfeffer thought, of exclusively individually acquired dis-
positions, may be proved by the following experiments : —
Seedlings of the Acacia, Albizzia lophanta, kept hitherto
in the dark, were subjected by me to intermittent artificial
light and darkness, a cycle of six hours being chosen in
one series of cases, and a cycle of twenty-four hours
in another series. When, after several weeks' exposure
to this kind of thing, intermittent illumination was
discontinued and the plants were left altogether either in
light or in darkness, they continued for some time their
movements of leaves, but not in the cycle of six and
twenty-four hours respectively which I had tried to
induce, but at intervals of twelve hours ; from which it
may be gathered that the tendency to a cycle of twelve
hours is an inherited one. Seedlings, which from the
beginning were left either in complete darkness or in
continual light, did not show periodic movements at
all. Those in the dark kept their leaves closed, while
those constantly exposed to the light took up and main-
tained with unfolded leaves an angular position, which
varied in different individuals from 135° to 180°. From
these observations, we conclude that periodic illumination
and obscuration contribute to the ecphory of engrams
whose corresponding reaction manifests itself in regular
leaf-movement every twelve hours. The periodic change

74 THE MNEME

in illumination is necessary for the ecphory of the engram.
But that the exposure to light itself has not generated
the engram, that is, the tendency to move in cycles of
twelve hours, may be gathered from the fact that after
the cessation of the intermittent illumination, the opening
and closing of the leaves do not take place in cycles of
six and twenty-four hours, the experimental periods, but
in one of twelve hours, a period to which the individual
under observation has never been exposed, but which
its ancestors for many generations have experienced,
and which may, therefore, be regarded as inherited.

Further evidence of mnemic periodicity, the engraphic
factors of which are readily recognisable, may be derived
from the work of Bohn, Schleip and others. A diurnal
period of an undoubtedly mnemic character may be
observed in Crustaceae. In the case of a shrimp, Hippolyte
varians, the periodicity has been thoroughly investigated.
Certain observations of Gamble and Keeble on Palcemon
squilla suggest that among the Crustaceae there has been
an hereditary fixing of this periodicity. Of many other
animals it might be said that they show a mnemic diurnal
periodicity, that is one which persists after the discon-
tinuance of the periodical exposure to light. But it
has not yet been determined in these cases whether the
periodicity has acquired an hereditary character. Special
mention may here be made of Bohn's observations on
Actiniae, Worms, and Molluscae, and of Schleip's recently
published investigations on the " Sceptre " locust, Dixippus
morosus. Schleip gives a good index to the literature of
the subject.

The diurnal periodicity, however, is but a part of the
mnemic periodicity. Bohn showed, for instance, that
the alternation of tidal ebb and flow produced a six
hours' cycle in the shore-dwelling Actini, Turbellaria, and
Snails.

Many pertinent observations on this subject might be
quoted, but a comprehensive treatment of the problem
of the engraphic, and of the hereditarily engraphic deter-
mination of periodic stimulation, will be given later.

ENGRAPHIC ACTION ON PROGENY 75

In the course of the preceding chapter we have traced
phasogeneous ecphory amongst the various ecphoric
influences (p. 56), and described how, when a definite
phase of development is reached, a state of the irritable
substance is produced which acts ecphorically on a definite
engram. Inherited dispositions, becoming manifest by
phasogeneous ecphory, may be met with in very great
numbers in plant and animal organisms. To mention
but one case : — The segmentation of the ovum of Synapta
digitata, an echinoderm, occurs, according to Selenka,
in nine phases ; the co-existing cells dividing equally
in nine successive stages. When the cells number 512,
a process regularly takes place, which is usually de-
scribed as " gastrulation." A rapid cell-multipUcation
by division begins at one pole of the ovum, and
this part invaginates into the cleavage cavity. Simul-
taneously all the cells develop cilia on their exterior,
and the organism begins to rotate within the membrane
of the ovum. May we not, therefore, say in the case
of Synapta that the reaching of the " 512 cell stage,"
or, as we may phrase it, the energetic condition of the
"512 cell stage," acts ecphorically on an inherited disposi-
tion which becomes manifest to us in a number of plastic
and motor reactions such as gastrulation, cilia formation,
and ciliary motion ?

We must bear in mind, however, that the description
" phasogeneous ecphory " covers a state which in each
individual case is open to a more exact analysis. By
the specific development-phase of an organism we
understand its entire morphological and physiological
state at any given moment. In members of the same
species the development of the organism occurs with a
certain regularity, but by no means in an absolutely
identical manner. The corresponding phases, or stages,
or total states are similar but not identical, even in twins.
We shall, therefore, find it impossible to define in a
universally acceptable manner a total phase other than
that of the concrete case with which we are at any one
moment concerned. However, no serious difficulty arises

76 THE MNEME

from that as regards our conception of phasogeneous
ecphory. For it is conveniently characteristic of each
ecphory that a partial return of a definite energetic
condition suiifices to awaken the engram from its latent
state. A new internal energetic condition marks each
entry into a new phase of development. Slight deviations
in the developmental phase will, in the greater number
of cases, be practically of no consequence so far as the
ecphoric action is concerned, seeing that for the ecphory
the partial return of the energetic condition is sufficient.

This general statement, of course, is not to be regarded
as the final word. The great symphony of single com-
ponents, which each organic development represents,
contains, in addition to that general dependence of each
successive total phase on its predecessor, numerous specific
dependencies within the separate developing organic
systems.

A simile may be useful. In the reproduction of a
piece of music, say a pianoforte sonata, the whole pro-
gresses from phase to phase, each new bar growing, so
to speak, out of its predecessor, and its mode of exit
determining the mode of entry of its successor. But, in
addition, there co-exist as many specific dependencies
within the single parts as there are such parts. Each of
these parts is closely knitted with the whole, but it also
progresses with a certain independence within its own
succession ; and in the case of a not very skilled pianist
it may happen that the bar he is playing with his right
hand may finish slightly in advance of the bar played
with his left hand, as, for instance, when he has to play
one rhythm in triple time with his right hand, and another
in common time with his left. Nevertheless, both parts,
although progressing independently, will coincide again
either at the conclusion of the movement or at the entry
of the new phase — in this latter case, a new bar. So also
in the wonderful symphony of an organic development
there exist, besides the general relationship of succeed-
ing simultaneous complexes, numerous specific relations,
which again are subordinated to the whole. Within the

ENGRAPHIC ACTION ON PROGENY 77

separate cycles, slight oscillations in the ensemble may
occur, corresponding to what is called "Heterochrony"
in embryology. Here, also, the common association of
the single components within the simultaneous complexes
usually forces the vagaries of specific cycles back into
the general rhythm. In the second part of this work
(p. 96) we shall have an opportunity of a closer con-
sideration of the more or less intimate relations existing
between the single components of successive simultaneous
complexes.

From the point of view just enunciated, it is incumbent
upon us to search within the total phases for specific
changes, which are either connected with specific ecphories
or indicative of ecphoric manifestations. Thus, in the
above quoted case of the gastrulation of a definite Echino-
derm ovum, reaction took place on the attainment of
the " 512 cell stage." Is it now possible to trace within
the total state of the phase those specific appearances
on which the ecphory depends ? The observations and
conclusions of a large number of scientific workers, among
whom special mention may be made of Th. H. Morgan,
Driesch, and Boveri, suggest the answer. In inducing
segmentation in fragments of the Echinoderm eggs, as
also in isolated blastomeres, gastrulation resulted, not
simply after the characteristic numerical segmentation
phase of the normal egg had been reached, but after
the cells had been reduced to a certain size by continued
subdivision. The true ecphorising factor here was,
therefore, not simply the completion of a certain number
of cell-divisions and the attainment of a definite number
of cells, but primarily the attainment of a certain size of
cell in division. The creation of a definite relation,
variable within certain limits, between the chromatin
mass and the protoplasmic mass within the cell was
necessary to the ecphory.

Yet another example more exactly defining the most
important element in specific phasogeneous ecphory may
be given. When a vertebrate embryo has reached a
certain stage of development, the formation of the lens

78 THE MNEME

takes place. Herbst and Speman, independently of
each other, have demonstrated that the formation of
the lens does not follow on the mere realisation of one or
the other total phase, but specifically is dependent on
the stimulus exercised by the retinal layer of the optic
vesicle when it comes into close contact with the epidermis.

Speman was able to furnish experimental evidence
that in the case of the common frog {Rana fusca) no
lens formation takes place unless the optic vesicle touches
the epidermis. The moment, however, the optic vesicle
comes into contact with the epidermis, lens growth begins
at the point of contact. Otherwise the epidermis retains
its dark pigmentation, and the growth of transparent
corneal epithelium is impossible.

With these facts before us it would seem necessary, at
least in this case, to narrow the cbnception of phasogeneous
ecphory, and to make it dependent on a single element,
namely, contact-stimialus exercised by the retinal layer
of the optic vesicle on the epidermis. It has, however,
been discovered since the work of Herbst and Speman
that there are cases where the lens formation does take
place on the mere entry into a total phase, even though
the main component, the contact-stimulus of the optic
vesicle on the epidermis, be lacking in this state of the
energetic condition. Mencl, in his work on the embryo
of the salmon, found that lens formation can take place
in the entire absence of the optic vesicle. Speman himself
admitted that while in the case of the common frog
(Rana fusca), the toad (Bombinator igneus), and probably
also Triton tceniatus, contact-stimulus seems to be indis-
pensable, in the case of the edible frog {Rana esculenta)
lens formation takes place regularly without this contact,
that is, as soon as the respective total phase has been
reached.

The American Rana palustris seems to resemble in this
respect Rana esculenta, according to experiments by King.
It is interesting in this connection to note that a small
fragment of the optic vesicle in Bombinator suffices to
start lens formation.

ENGRAPHIC ACTION ON PROGENY 79

Speaking generally, we may say in regard to this
problem that there is a developmentaJ component,
namely, contact-stimulus of the optic vesicle on the
epidermis, which exercises a strong ecphoric action on
the engram complex, and that the ecphory majiifests
itself in the reaction of lens formation. As first shown
by Lewis in his work on Rana silvaiica, this contact-
stimulus of the optic vesicle on the skin can also
ecphorise lens formation on parts of the epidermis other
than those which the normal development demands.

He removed the optic vesicle at a very early stage from
the brain, and inserted it beneath the epidermis of some
other part of the body, such as the abdomen, etc. In
numerous cases he succeeded in obtaining the formation
of a lens in parts of the body where normally this never
occurs. Speman, after repeated experiments on the
toad Bombinator, considers it probable that, if not the
skin of the abdomen or of the trunk, at any rate that
of the head behind the eye, is able to form a lens on
contact with the optic vesicle. Recent investigations
by G. Ekman have even shown that during a definite
period the entire ectoderm of the green frog {Hyla
. arhorea), with the possible exception of the ear and
nose primordia, is capable of lens formation. In com-
parison with the ecphoric action of the stimulus, the
other components of the energetic conditions of that
phase play but a minor part in their ecphoric action
on this engram complex ; they are, however, not
altogether without effect, as may be demoijstrated in
the cases of Rana esculenta and Rana palustris, where
they suffice for ecphory on elimination of such contact-
stimulus. In our conception of the process as one of
phasogeneous ecphory the apparent irregularities and
exceptions, which otherwise would loom large, disappear.
Indeed, the very facts just given — material available
since the first edition of The Mneme — show us in the
clearest manner that an understanding of these pro-
cesses can be gained only in the light of the theory of
ecphory.

8o THE MNEME

From what has been said we reach this important
conclusion. In the irritable substance of protists, of
animals, and of plants are properties, "excitation-dispo-
sitions," which as a rule are characterised by latency.
Like the individually acquired engrams, they are evoked
from this state of latency by stimuli, on the cessation of
which they sooner or later lapse into latency. Each
recurrence of the ecphoric influence effects the return
of the respective state of excitement, and this becomes
manifest to us by its own proper reactions.

What finally determines our conclusion that these are
inherited engrams is the nature of the influences calling
them into activity. They bear, as we have briefly
demonstrated, the definite character of an ecphory in
part direct, in part associative, or chronogeneous, or
phasogeneous. Finally, we note that these inherited
excitation-dispositions behave in every respect like engrams.
Only their origin, not their nature, is problematical.

We have still to deal with the problem of differentiating
the characteristics of those inherited excitation-dispositionSj
which in our view are to be regarded as engrams, from
those of other inherited dispositions which have not
so to be regarded. It is natural enough to look for dis-
tinguishing marks in those characteristic properties of
the inherited engrams which are typical also of the engram
as such. The engram, we may repeat, is characterised,
first, by a latent condition ; secondly, by the fact that
each recurrence of the liberating influence effects the
return of the corresponding state of excitement ; thirdly,
by the ecphoric character of the liberating influences ; and
finally, by its power, through addition of new engraphic
influences, to affect the function of the inherited dispositions.

According to their possession or non-possession of
these characteristics, the inherited excitation-dispositions
may or may not be regarded as inherited engrams.

The first characteristic, a condition of latency, is
useless for the purposes of differentiation, simply because
in this discussion we have regarded all properties of the
organic substance as dispositions or predispositions. The

ENGRAPHIC ACTION ON PROGENY 8i

whole problem is concerned with dispositions which are
generally latent.

The second characteristic is that with each recurrence
of the hberating influence there is a return of the
corresponding state of excitement. Are there now any
inherited dispositions in which this is not the case ?
Personally, I am unaware of any dispositions of the
organic irritable substance which, without the occurrence
of other changes of state — for example, the entry into a
totally different stage of development — ^have exhausted
themselves by having been repeatedly roused into their
corresponding states of excitement. But too frequent
repetition and too long duration of the excitation may
produce a state of fatigue, leading to the enfeeblement
of the excitation and of the reactions caused by it.
If, however, sufficient time is given for the organism
to recover, no exhaustion takes place, or, at any rate,
not while it remains in a state of primitive energy ; but
on the contrary, an increase in the predisposition to
react usually sets in. Many dispositions are only elicited
once in the individual life of the organism, as witness
the lapse of the different development phases in onto-
genesis, and of instincts which normally become manifest
only once in the individual Hfe. In these cases it is not
that the disposition has been exhausted, but only that
the end has been reached of the situation which acts
ecphorically on the disposition ; for by artificiEdly throwing
back an organism to a state through which it has already
passed, the disposition can be roused again. Mention
need only be made of the numerous cases of regeneration in
embryos and fully developed animals in order to show
that the disposition, normally evoked once only in the
individual life, is not thereby exhausted. This is the
case, too, with those excitation-dispositions, the reactions
of which belong to the motor or secretive area. We
have already on page 71 related how caterpillars, which
normally make only one cocoon in their lifetime, can
be induced to spin parts, or the whole, over and over
again.

6

82 THE MNEME

The latent condition of inherited dispositions can
readily be inferred. Further, such dispositions are
characterised by a persistent vitality ; for far from being
enfeebled or slowly exhausted by repeatedly being roused
into activity, they, on the contrary, with proper regard
to time for recovery are strengthened by it.

Perhaps, in dealing with the third characteristic of
an engram, we may succeed in tracing inherited dispositions
which differ in their response to ecphoric influences from
those whose activity cannot be traced to such ecphory.
An exact determination of the difficulty is impossible,
for a liberating influence can only be called ecphoric with
certainty, provided we know the engraphic stimulus,
and are thus in a position to distinguish by comparison
between the ecphoric and the engraphic influences. Our
real difficulty lies in the fact that, in the great majority of
inherited dispositions, the engraphic stimuli are unknown
to us, and that in cases of immediate manifestations we
are able only under especially favourable conditions
to recognise their ecphoric character. Reference may
be made to the elaborate pantomime of a bath without
water that a young magpie was induced to go through
by the contact of its beak with water. On the other
hand, we are not justified in eliminating cases where
the ecphoric character of the actuating factor is less
apparent, or in denying the engram-nature of the respective
reactions.

Amongst inherited dispositions the irritability as such
is not, of course, to be included, for that is the indis-
pensable condition for the acquisition of engrams. But the
specific evolution of this irritability is brought about by
its one fundamental property of susceptibility to engraphic
influences. The irritability, as it presents itself to us
to-day in the single organism after a history of many
millions of years, is charged with innumerable engrams,
whose presence renders the irritability an exceedingly
complex thing.

Earlier in our investigation, when we defined the primary
state of indifference, we said that we wished thereby

EN GRAPHIC ACTION ON PROGENY 83

simply to imply the state of the organism at the beginning
of our various observations and experiments. Even
if. our subject is an organism just separated from its
mother, it will only be a blank sheet in respect to its
individual mnerae. If we submit such an organism
for the first time to a stimulus and describe the result
as a simple synchronous stimulation, the stimulus can
only be regarded relatively as original. In the majority
of such instances it will be partly a case of the ecphory
of inherited engrams, or it may be that such ecphories
will mingle with the action on the primary, so to say,
pre-engraphic irritabihty. We are, therefore, unable to
eliminate with absolute certainty the ecphoric co-operation
in any stimulation. If we were in a position to examine
organic matter newly produced by spontaneous generation,
then, and only then, should we be able to observe purely
synchronous stimulation without any trace of ecphory.
This diificulty will also prevent us, in deahng with inherited
dispositions, from distinguishing by the ecphoric or non-
ecphoric character of the actuating influences the engraphic
from the non-engraphic.

Finally, we turn to the consideration of the fourth
characteristic of an engram, and enquire as to the kind
of inherited disposition which can be influenced by newly
generated engrams. The mere possibility of influencing
such a disposition engraphically is sufficient justification
for regarding it as an inherited engram. But the im-
possibihty of so doing is no sure argument in favour of
the opposite conclusion, simply because our failure to
influence the disposition engraphically is partly caused
by the imperfection of our experiments ; some new experi-
mental method may prove the possibihty of influence.
The degree of difficulty in influencing inherited dispositions
engraphically varies greatly. For example, those inherited
dispositions of the higher animals, for the ecphory of
which a definite speciaUsation of parts of the central
nervous system is required, are more easily influenced
engraphically than less speciaUsed dispositions. But,
indeed. I do not know of any class of inherited dispositions

84 THE MNEME

which is entirely exempt from engraphic influence, nor
of any group of organisms which possess rigid, non-
modifiable, inherited dispositions. Among the Bacteria,
many inherited dispositions can readily be affected by
newly acquired engrams ; with comparative ease it is
even possible to secure the hereditary continuation
of the engraphic influence. So also with many other
inherited dispositions, such as the heliotropism of uni-
cellular organisms like the Flagellates, and the tropisms,
periodic movements, rate of growth, and many other
characteristics of Plants.

In a limited but fairly imposing number of cases we are
able to influence inherited dispositions so effectively that
the newly added engram not only remains in force during
the individual life of the organism, but is transmitted
to the offspring. Attention is directed to those cited on
pages 57-64. The cases are limited partly by reason
of the imperfection and too brief duration of our experi-
ments, and also because of the short period during which
the germ-cells of many, perhaps of all, organisms are to any
considerable extent sensitive to any engraphic influence.
From our prescribed human point of view the hereditary en-
graphic variability might almost be described as capricious.

Thus it appears impossible to me to divide inherited
dispositions into two categories, namely, those that are
to be regarded as engrams, and those that are not. If,
therefore, we look upon each specifically evolved form
of irritability as engraphically complicated, we have to
prove the value of the conception, and ask whether it
is possible to apply it logically in concrete cases. Does
it help us to a " complete description in the simplest
manner," to quote Kirchoff's criterion ? If we can
furnish proof that facts do not contradict our idea of
the engram-nature of inherited dispositions, and also
that such a conception throws new light on this aspect of
organic phenomena, we have added a new and important
link to the chain of probability we are endeavouring to
forge.

This part of our task will be attempted in the third

EN GRAPHIC ACTION ON PROGENY 85

part of this book, where the action of mnemic processes
in Ontogenesis will be investigated, and where it will
be shown how the enigmatic phenomena of the normal
ontogenetic processes and of those modified by interfer-
ences, as well as the phenomena of regeneration, are
brought within closer reach of our understanding by' our
conception of them as a function of the Mneme. Before
we elaborate this phase of our subject, we shall in the
second part of this work deepen and extend our founda-
tions by subjecting the mnemic fundamental phenomena
to a systematic investigation.

POSTCRIPT TO THE THIRD GERMAN EDITION

Since the appearance of the first edition in 1904, the
material from which evidence in support of the engraphic
origin of dispositions acquired in the long process of
evolution can be drawn has been extraordinarily in-
creased. New experiments by Blaringhem, Klebs, Bordage,
Kammerer, Pictet, Przibram, and Summer, to name but
the principal workers in this field, are in entire harmony
with the earlier investigations of Chauvin, Standfuss,
Fischer, Schroder, and others. It has been most clearly
demonstrated that, with appropriate experimental arrange-
ments, dispositions of all kinds can be created de novo ;
that changes in the manner of manifestation of the already
existing dispositions can be obtained without difficulty ;
and that the new acquisitions of the organism, of what-
ever origin they may be, present themselves without
exception as products of stimulation or induction.
They are to be regarded as engrams, and we can prove
their hereditary nature, if in appropriate circumstances
they are generated during the sensitive period of the
germ-cells. They also behave like dispositions acquired
during evolution, for, as the crossing experiments of
Tower and Kammerer show, they can under suitable con-
ditions be transmitted according to Mendel's principles of
segregation. On the strength of these numerous indis-

86 THE MNEME

putable agreements, the conclusion seems fully justified
that dispositions acquired during evolution also have
arisen engraphically.

I have just completed the compilation of relevant
data in a special work entitled. The Problem of the Inherit-
ance of Acquired Characters {Das Problem Der Vererbung
Erworbener Eigenschaften, Wilhelm Engelmann, Leipzig,
1912).

As shown in that book, experimental research has
proved that all new acquisitions of the organism must
be regarded as products of stimulation or induction. We
are, therefore, driven to elaborate some such Engram
Theory as the one with which we are now concerned.
In the following sections of this book it will be our task
to show how our general understanding of organic pro-
cesses will profit by the working out of this theory.

PART II

CHAPTER IV

THE MUTUAL RELATIONS OF ENGRAMS :
SIMULTANEOUS AND SUCCESSIVE
ASSOCIATION

In the introductory part we discussed the genesis of
engrams and the various phases of their existence. From
an analysis of organic states we formulated the following
definition : — " An engram is the result of engraphic action,
and implies an altered disposition of the irritable sub-
stance towards a recurrence of the state of excitement
produced by the original stimulus. The organic substance
so affected by engraphic action shows itself specifically
predisposed to the state of excitement induced by the
original stimulus. On being subjected again to this
stimulus, or to other influences, the basis of which is
invariably the partial recurrence of a definite energetic
condition, the original state of excitement is reproduced."
With this definition in mind, we have next to trace
more precisely the relations of the various engrams
generated and conserved in the same organism. In the
nature of things, the organism can very rarely be affected
and influenced by a single stimulus alone, as for instance,
by a single light-ray of a definite wave length. The
stimulus, even when belonging to a distinct category
such as the photic stimuli, is invariably of a complex
nature. We need only think of the various elements
combined in the visual impressions or sounds which act
upon our organism. When we see a landscape, or when
we listen to a piece of music, our organism is in both
cases simultaneously excited by a large number of stimuli,

8,

go THE MNEME

which, however, do not blend into something homo-
geneous, but group themselves in juxtaposition. Thus,
we are conscious of the juxtaposition of the photic stimuli
which impinge on our retina, and such juxtaposition
we call a picture. In similar fashion we term simul-
taneous polyphonic acoustic stimuli a chord or a discord.
We do not attempt an explanation of this fact, but accept
it as such. In keeping with this co-ordinated reception
by the organism of simultaneously acting stimuli, which
we may describe as a co-ordinated synchronous effect of
stimulation, the engraphic effect of this stimulation is
also a co-ordinated one. This means that subsequently,
on the ecphory of the respective engrams, a state of excite-
ment arises which corresponds to the co-ordinated effect
of those stimuli which previously acted synchronically in
juxtaposition.

This statement, which at first sight appears to be only
a roundabout way of saying that co-acting stimuli are
received by the organism and are mnemically reproduced
in juxtaposition, involves far-reaching consequences, and
helps us to a surprising perception of the deeper connec-
tion of various data which so far we have obtained by
the analytical method alone. We have already noted
how, on the action of a complex stimulus of a definite
kind, a juxtaposition, not a mere diffuse blending, is
secured. A co-ordinated synchronous effect of the stimu-
lation also follows on the simultaneous action of various
stimuli, not only of the same, but also of different cate-
gories.

When an express train thunders past us rapidly, we
feel clearly the juxtaposition of visual and auditory
impressions, and we are able to reproduce this juxta-
position mnemically after the cessation of the stimuli,
if the latter have been sufficiently strong to act en-
graphically. It in no way affects the problem occup3^ng
us here that, at the reception of stimuli belonging to
different categories, different receptor-organs are called
into action. Co-ordinated vision, for example, is con-
ditioned by the excitation of numerous specific receptor-

THE MUTUAL RELATIONS OF ENGRAMS 91

elements, such as the cones and rods of the retina. Tactual
and auditory impressions involve like conditions. What
interests us here is the fact that the organism as a whole
is able to perceive simultaneously different stimuli in
juxtaposition, and always does so perceive them ; and
that the effect of these involves elements contiguous
and disparate which refuse to blend into something
homogeneous. Thus, at any given moment a firmly
connected, co-ordinated total of excitations exists in each
individual, and this is manifest for Self in the conscious-
ness of the juxtaposition of sensations ; and in the case
of other organisms by numerous objectively perceptible
reactions. We will call this totality the simultaneous
excitation-complex.

From the coherent whole of this simultaneous complex
we may, for the sake of theory, abstract the action of
a single stimulus or stimulus-category, and thereby
faciHtate our understanding of this multitudinous com-
plex of simultaneous excitations. But we must remember
that in so doing we strain the actual facts of observa-
tion and arbitrarily undo the nexus of the whole. This
must be borne in mind when we consider not only the
synchronous but also the engraphic effect of the stimuli.
The individual stimulus does not liberate one isolated
synchronous excitation and make an isolated engram.
Such detachment is well-nigh impossible in Nature. What ^
we find is this : that a simultaneous excitation-complex
as such is, after the lapse of the synchronous excitations,
engraphically fixed in its totality. That which remains
is a simultaneous engram-complex.

Let us examine more closely the engraphic action which
the simultaneous excitation-complex exercises on the
organism. We may characterise this simultaneous complex
as the product of the excitations resulting from the entire
energetic condition. By " energetic condition " we mean
not only the external energies affecting the organism, but
also its internal energetic state. The latter is at least
as important as the forces acting on it from without —
a fact which becomes evident on subjecting the organism

92 THE MNEME

of a higher animal to the influence of the same external
stimuU during sleep and in the waking state. In the
waking state the internal energetic state of an organism
varies greatly at different times, so that the same external
energetic condition produces at various moments alto-
gether different simultaneous excitation-complexes ; the
subsequent engraphic effect varies accordingly, different
engrams corresponding to different excitation-complexes.

As the mnemic reproduction is usually weaker than
the original excitations were at the time when they
created the engrams, we must not be surprised that on
the ecphory of the engram-complex many of the com-
ponents of the complex do not become manifest at all,,
that is, they cannot be recognised by reactions. Thus
it seems that of successive experiences a few engrams
only are conceived. But as already stated, most of
these engrams are not simple units, but products of highly
composite excitation-complexes.

Because only a small fraction of a simultaneous excita-
tion-complex manifests itself clearly on mnemic repro-
duction, we get the impression, not of the recurrence of
a former complete energetic condition, but only of a
specific section. We were once standing by the Bay
of Naples and saw Capri lying before us ; near by an
organ-grinder played on a large barrel-organ ; a peculiar
smell of oil reached us from a neighbouring " trattoria " ;
the sun was beating pitilessly on our backs ; and our
boots, in which we had been tramping about for hours,
pinched us. Many years after, a similar smell of oil
ecphorised most vividly the optic engram of Capri, and
even now this smell has invariably the same effect. The
other elements of the simultaneous situation are not
ecphorised by the smell of oil. The melody of the barrel-
organ, the heat of the sun, the discomfort of the boots
are ecphorised neither by the smell of oil nor by the
renewed experience of Capri ; and on their recurrence as
original stimuli they fail to effect the ecphory of those
two engram-complexes, but this is no evidence that they
have not acted engraphically at all It may be that

THE MUTUAL RELATIONS OF ENGRAMS 93

a friend reminds us of the sufferings we then endured
from the heat and our tight boots, or he plays the organ
melody on the piano to us, and we then find that these
elements of the energetic condition have also acted en-
graphically, but that for their ecphory some special kind
of assistance is necessary.

The case illustrates the fact that the stimulus of which
we become most vividly conscious has not always the
most powerful engraphic effect. The pressure of the
boots was in its way more insistent than the smell from
the oil ; yet it was the smell-element of the excitation-
complex which apparently attained the stronger engraphic
position. We need not enter more closely into a dis-
cussion of these specific questions, but it is well to em-
phasise that, as a general rule, the nature of the stimulus
and the momentary state of the organism are together
the principal factors in determining the intensity of the
engraphic influence, and that according to circumstances
the influence of the one or other factor predominates.

In the many simultaneous engram-complexes which"'
we retain, we often find that, together with important
impressions, there are many commonplace indifferent ones
mnemically fixed. One of my earliest recollections is
of a garden at. Kreuznach, in which as a child of three
years of age I had been stung by a wasp. Even to-day
I could draw the shape of the flower-beds and the dis-
tribution of the rose-trees and my relative position in
the garden at the moment I was stung. "^

Darwin tells us in his autobiography that the solution
of an important problem, which furnished him with the
key to many hitherto enigmatic cases, came quite sud-
denly to him during a drive. He adds the remark : "I
can remember the very spot in the road, whilst in my
carriage, when to my joy the solution occurred to me,
and this was long after I had come to Down." It would
be easy to quote many more observations which show
that in moments of intense emotion, besides the main
impressions, relatively unimportant parts of the simul-
taneous excitation-complex also act engraphically in a

94 THE MNEME

Surprisingly strong manner. Later, they naturally show
themselves indissolubly associated with the main impres-
sions and accordingly act ecphorically on these.

I think our investigations have shown that the facts
of simultaneous association follow from the one funda-
mental postulate that the organic substance responds to a
number of simultaneous stimuli with an orderly, coherent
juxtaposition of excitations, and that this simultaneous
excitation-complex acts engraphically as a complex.

The single simultaneous excitation-complexes are re-
y lated to each other in a temporal succession marked by
specific characteristics due to their varied mode of origin.
The succession is cdntinuous ; that is, one complex
^ merges into the next. A break, in the sense of successive
complexes separated by a section devoid of excitations,
never occurs, not even in sleep or in those periods of rest
when there is a discontinuity of consciousness.

Further, the relation of complexes is unilinear. Each
simultaneous complex is connected with but two others,
its immediate predecessor and its immediate successor.
As no complex exactly resembles another, although a
periodic recurrence of single components of the com-
plexes is characteristic of many life processes, the sequence
of these complexes is non-reversible ; consequently it
makes a fundamental difference whether we regard them
in the order of their genesis, i.e. from the earlier to the
later, or vice versa, from the later to the earlier

Just as the simultaneous engram appears on ecphory
as the faithful, if somewhat weakened, image of the
simultaneous excitation-complex to which it owes its
origin, showing, therefore, the same co-ordinated juxta-
position of the various stimulations, so in a sequence of
^ engram-complexes the same unbroken, unilinear, pro-
gressive arrangement, in which the original excitation-
complexes were grouped, reappears on each ecphory of
the sequence, proving thereby that the succession is
engraphically registered.

We have noted that each separate simultaneous en-
gram complex presents a co-ordinated juxtaposition of the

THE MUTUAL RELATIONS OF ENGRAMS 95

various stimulations, and that it is followed immediately
by a like co-ordinated complex.

Hereby a situation is created in which each fresh
engram has to be allocated to a definite place. From the
very beginning, therefore, the engram stands in closer
relation to some, and in more distant relation to other,
simultaneous engrams.

It follows from this, that within any simultaneous
complex there are components already more closely
associated with some fellow components than with others,
and that this difference of relationship also obtains in the
case of successive associations.

In certain circumstances, close associations between
engrams may be created which the relative position in the
simultaneous complex would not lead us to expect. But
this follows only from the specific strength of engraphic
action exercised on the respective engrams at their genesis,
to the exclusion of their fellow components.

Strength of stimuU, frequency of the simultaneous ^
stimulation, the focussing of attention effect the close
association of entirely heterogeneous engrams. So a
smell of oil becomes more closely connected with the
visual image of Capri than with other components of
the same simultaneous complex.

Within any simultaneous engram-complex, as well as
within successively associated complexes, a more intimate ^
association exists between engrams of like stimiilus-
quaUty than between engrams of different stimulus-
quality. Further, given the same stimulus quahty,
engrams of nearly related origin are more closely asso-
ciated than those of more distant relationship.

The closer or more distant relationship of the engrams "
refers, of course, to their mutual ecphoric action. A
few examples may, however, make the meaning clearer.
The execution of an impressive dance to an easily re-
membered tune generates a double succession of acoustic
and optic engrams which, series by series, are easily
ecphorised. Generally, one series acts ecphorically on
the other. On hearing the tune, I have the visual image

96 THE MNEME

of the dancer, or on seeing the dancer, I have the auditory
image of the melody. But in the lack of an often repeated
experience of dance and tune together, the optic engram
does not attain that close association with the acoustic
engram whereby the simultaneous movement of the dancer
corresponds exactly to each bar of the melody.

Again, in engram-complexes originated from one and
the same stimulus-quality, we may discern the existence
of more intimate connections between some than between
other components of the successions. Any one of fair
musical abiUty, to whom a polyphonous piece of music
has often been played or sung, is able after a little time
to reproduce as a memory-image, or to sing, or to play
the sequence of each single part.

The performance does not demand a highly specialised
musical training, as a child can easily reproduce the
sequences in the case of a two-part song. But the ability
'' to reproduce in that way can only be explained on the
assumption that the tone sequences within each particiolar
part are more closely associated with each other than are
the sequences of one part with those of another. Should
this not be the case, the reproducer in his attempt to
capture a single part would helplessly vacillate between
the collateral progressions, especially if he had set himself
to render a part other than the dominant.

In the following diagram we trace a case, very much
simpUfied for our purpose, in which the relations of the
acoustic components have been worked out. The con-
nections between the four successive engram-complexes
can readily be seen.

We have now reached the following position : —

The external and internal stimuli acting simultaneously
/and successively on the organism generate in it excitation-
complexes which represent an orderly juxtaposition in
their simultaneous action, and an unbroken, unilinegr, and
iiion-reversible sequence in their succession.

^fhe appeafance"6F" these excitation-complexes implies
a permanent affection of the organism. This we charac-
terised as engraphic, meaning thereby the creation of

V

n<'

JittuAJUi/tieevi

' \

■V-

\

1

\

\i 1

\

V

' "^

4

1

1

1

1
1

T

1

V.

1- *^ 1 -^

j,<^

1

Mn^J£,£/^iafV^eii^

K

1 H

k "

^ t ^

r ~

r

\

r /

^ ^

\

, .¥

1
1
1
1

1
1
1
1

1

V

1- -i i -^

L.'-

1

d^ntM£^a>t%eou4

I^ V

J 1 ■ ■ VJ

^

H

^ 1

' 1

\

'

1

1
1
1

1

1
1

■ J

|«. rt 1 «f)

-n^-

-Tti-

M/)r\4A/tfO/n't/>ujL

J

1 '^ ^

k

r

Yoiae. %t^

45 "Kc'^Jro >

J

V

1
1

^
^

Sit

98 THE MNEME

an increased disposition of the organism to reproduce
those excitation-complexes. It is obvious, therefore, that
things recur as on their first appearance, namely as an
orderly juxtaposition, and an unbroken, unilinear, and
non-reversible sequence.

It would be the exception to this that would naturally
call for a special explanation.

We described the co-ordination of the simultaneous
excitation-complexes, and the subsequent engram-com-
plexes as unbroken, unilinear, and non-reversible. The
expression " unbroken " implies the continuity in which
no division can be observed, division being solely a function
of our intellect.

This becomes clear at once when we try to understand
what we mean in this connection by the idea of division.
In speaking of simultaneous excitation, that is, of excita-
tions existing at the same time, the answer to the question
regarding the duration of such simultaneity can only
be " infinitesimally short," in view of the enormous
number of influences acting on the organism at any one
moment.

As each excitation of the organic substance possesses
a measurable duration, and consequently is never in-
finitesimally small, we have, at the juxtaposition of
simultaneous infinitesimally brief states, to apply to
the objects a purely logical principle of division db
extra.

The division is imperative and indispensable, for it is
in the very nature of our understanding that knowledge
of the world of phenomena proceeds by the aid of division
and rearrangement.

By regarding the excitations of an organism in point
of time as simultaneous and successive states, we obtain
an arrangement objectively justified, although based on an
arbitrary and purely logical division. In our examina-
tion of two successive simultaneous complexes, I and II,
we may find that regarding some components they differ
essentially from each other. The components a, h, c of
J may hiave disappeared. On the other hand, the com-

THE MUTUAL RELATIONS OF ENGRAMS 99

ponents x, y, z of II may have sprung into existence.
A number of components, however, will be regularly
carried forward from I to II, even if the external energetic
condition is suddenly and completely changed. For
example, I make a dive out of the bright hot sun-
shine into the cool dark depths of the river. First,
we note that a change of the external energetic con-
dition of my entire organism has taken place, not in an
infinitesimally small, but in a measurable, degree ; secondly,
that despite the change, many components of the
internal energetic condition of the organism remain
unaltered.

It foUows from these considerations that the simul-
taneous engram-complexes also represent theoretical and ^
not actual unities, and that in reality they pass into each
other without a break, notwithstanding that some of
their components are detached. The successive association
is not limited to the continuously progressing components
in the Une of simultaneous complexes, but holds good
also in regard to those components which are detached.
This becomes immediately evident if one thinks of a
melodic succession of sounds, or of noises such as the
beats of a drum, or the trampling of feet, or of those
lists of grammatical exceptions which in their bareness
seem devoid of sense. A little consideration convinces >
us that one can hardly speak of an association in the
permanently continuous and, therefore, unchanging com-
ponents of the successive engram-complexes. It is only
the varying components of the successive complexes,
those that come and go, that allow us to apply the idea
of connection and association. To them we apply the
objective method, arguing from the appearance and dis-
appearance of objectively perceivable reactions to the
idea of association. The conception of connection and ^
association surely presupposes the discontinuity of what
has to be connected.

Direct observation teaches us that the engrams left ^
by the discontinuous components of successive excitations
are themselves sviccf§§iyely agsgci^ted, even when the

roo THE MNEME

separating interval is appreciably measurable. The en-
graphic fixation of a melody is as sure when the sounds
follow each other " staccato " with appreciable intervals as
when they are joined together " legato." But the intervals
'^ must not be too long if the engrams are to become suc-
cessively associated. Tone sequences with intervals of
several minutes find it almost impossible to establish
themselves engraphically. It will be the task of later
experimental research to define for each case the maxi-
mum length of interval at which a successive association
still takes place. What has been said regarding acoustic
engrams applies with equal force to all other engrams.
They must not be separated by too long intervals from
each other, if the repetition of the earlier has to act
engraphically on the later.

The reason for this limitation in time for the successive
association becomes clear to us, when we examine the
relation of the successive with the simultaneous associa-
tion, especially with regard to the possible derivation of
the former from the latter. The successive association
is indeed only a consequence of the simultaneous one.
As previously stated (p. 21), to each original stimulus
there corresponds a definite synchronous excitation. On
cessation of the stimulus the excitation does not suddenly
cease, but ebbs gradually with frequent, if faint, revivals,
before finally dying away. We described the excitation
in this phase of its disappearance as acoluthic exci-
tation. It must be pointed out, however, that the
excitation acts engraphically, not only during its
stronger S3mchronous, but also during its weaker acoluthic
phase.

It is true that in a succession of discontinuous stimuli
there can be no simultaneity of the synchronous excita-
tions. Let us call these C(syn), D(syn), E(syn), F(syn) —
but there does exist a simultaneity of the synchronous
excitation D(syn) with the vanishing acoluthic phase of
its predecessor — which may be denoted by ci(ac), c»(ac),
C3(ac), 0.
The following diagram shows the simultanepus relations

THE MUTUAL RELATIONS OF ENGRAMS loi

of the synchronous phase of an excitation with the aco-
luthic phases of its predecessors: —

Phase

1 2 8 4 S 6 ' 7

C(syii) — c»(ac) - c'(ac) - c3(ac) - o
1 1 1

1 1 1
D(syn) - i«(ac) - d'(ac} - d3(ac) - o
1 1 1

1 1 1
E(syn) - e»(ac) - e«(ac) - «J(ac) - o
1 1 1

1 1 1
F(syn) -/«(ac) -/«(ac) - /3(ac) -

We see here that simultaneity exists between the
synchronous excitation D(syn) and the acoluthic c'(ac) ;
between the synchronous E(syn), the acoluthic (i'(ac)
and the already weakened c»(ac) ; finally between the
synchronous F(syn) and the acoluthic «'(ac), the weaker
acoluthic d^{ac), and the vanishing c3{ac). This simul-
taneity of the excitations implies, of course, a simul-
taneous association of the engrams left behind by the
excitations, and thereby the possibility of reducing the
successive association to a simultaneous one of two
engrams, of which the first has been left by the synchronous
excitation, and the other by the acoluthic excitation of
its predecessor.

A glance at our diagram will also help us to under-
stand the important fact experimentally determined by
Ebbinghaus, that successive association not only exists
between members of a continuous series as from one
member to the next, but that each member is associated,
although less strongly, with its second following member,
and also, but in still weaker connection, with the third
following, and so on. A limit to the formation of such
associations is only set by the continuous diminution and
final complete extinction of the acoluthic excitations.
This also explains why a succession of excitations can
give rise to associated engrams only when, during the
original excitation, the time-intervals between the separate
members have been sufiiciently short to allow of the
connection of the acoluthic excitation of the preceding

102 THE MNEME

member with the synchronous excitation of its successor.
These ideas, Ughtly touched on here, are more precisely
formulated in the Mnemic Sensations.

Despite the derivation of the successive association

from the simultaneous one, the final result in each case

•^ is very different. This is shown in a striking manner

when we compare the ecphory of simultaneously asso-

^ ciated engrams with that of successively associated

engrams. True, in one respect congruity exists between

them. Also at_the ecphory of,5imultaneo usly associated

"" engrams, one must be ecphorised a little time before the

otherTi n ofg ^floecpfionsetfiii otEerT

"STmultaneously pfoHuced_ jengrams are link ed wit h
eguOTorce"iirboth directions, whilst_thi s is not the case
with succes siveTyproduced engrams, which constitutes a
difference of~Tundamentar importance. We arrive at
this statement from the fact that, if engram a is simul-
taneously associated with engram h, then under ordinary
circumstances the ecphorising power of a on 6 and of
6 on a is equal. But if a and b are successively asso-
ciated, the ecphory of a acts far more strongly ecphorically
on b, than the other way round.

This varying ecphoric strength in the successive asso-
ciation can be most easily demonstrated in dealing with
auditory engrams. Even the skilled musician may be
'' quite unable to recognise the most familiar tone passage
if it be played to him in reversed order. The same diffi-
culty applies in regard to the vocal sounds of which the
words of our language are composed.

When on my travels in Queensland I passed through the
squatter station Degilbo. Afterwards I heard its name
frequently mentioned, but I had no idea that the name
was simply the word " obliged " in reversed spelUng.
If chance had not enhghtened me, I should have regarded
this word as a typical native name. Again, it is not
possible to recite a poem in reversed sequence of words,
nor a meaningless combination of words like the lists of
gender, for example. If in a recitation we misplace the
V\ ^ words, this may only be evidence that the engrams have

THE MUTUAL RELATIONS OF ENGRAMS 103

been acquired in a false sequence, or that during the
learning of the poem we sought the assistance of simul-
taneous association. If, for example, a boy learning the
Latin rules of gender pictures to himself as an aid to
memory, while reciting the list " Men, people, rivers,
winds," how the men of a seafaring people enter the
mouth of a river under favourable winds, the normal
succession by the use of such simultaneous associations
becomes disturbed, because the components of the simul-
taneous association ecphorise each other indiscriminately.

We have seen that successively associated engrams act
ecphorically more strongly on each other in the order
of their genesis than conversely. But the evidence for ^
this rule becomes, as we have just seen, somewhat ob-
scured in the successions of auditory engrams by the
co-operation of simultaneous associations. This difficulty
arises oftener in cases of the succession of engrams which
have made their entrance through some other sense
receptors. Let us take the succession of visual impressions
during a walk. The succession of visual excitations,
and consequently the generation of successive visual
engrams, often occurs so slowly, and so many components
continually pass from the earUer into the later stages,
that the ecphoric predominance of the succeeding engram
over its predecessor in the chain might become almost
unrecognisable simply through the interplay of numerous
simultaneous connections.

That visual engrams in a succession act ecphorically "
with unequal strength may be adequately shown in the
following way: — The image of a movement is physio-
logically represented by a succession of visual excitations,
which, if of mnemic origin, represents a succession of
visual engrams. If the order of the movement be now
reversed — and this can easily be done by photographing
small phases of the movement, and then with the help
of either kinetoscope or cinematograph viewing these
photographs in the reversed way — we find that the
reversion affects us quite as strangely as would a reversed
tone-succession or the word " Degilbo " for " obUged."

104 THE MNEME

That is, its connection with the original movement is
not recognised at all.
Among human beings, engrams of the olfactory sense,
, and also of the gustatory sense, often enter into simul-
taneous association with engrams of other origin, but
very rarely with engrams of their own kind. Among
simultaneous olfactory excitations, the co-ordination is
not as definite and clear as among simultaneous visual,
auditory, or tactual excitations. Frequently, the olfactory
excitations so blend that they mutually neutralise each
other so far as their individual quality is concerned.
Nor do they form successive sequences of distinct units
as do tone-passages. Forel has rightly argued that this
defect can be chiefly attributed to the irregular manner
in which the stimuli enter our organism. The principal
cause of the inabiUty of the stimulus-successions to act
sufficiently clearly and quickly in the generation of
successively associated engrams lies in the nature of the
sense-organs, which serve as receptors of such stimuli.

In many insects, the engrams entering into the organism
through the receptor of the olfactory sense show them-
selves much more perfectly associated, simultaneously
and successively, than do the olfactory engrams in man.
Forel has convincingly argued that this distinction between
insects and man rests not so much on the greater or lesser
acumen of the senses as on the different mode by which
different organisms receive these stimulations. With the
ant the antennae are the sense-organs. By these the
ant cognises the objects in its way olfactorily. Through
such an organ it receives simultaneously and successively
associated excitations, and the immediately subsequent
corresponding engrams ; just as we human beings, and
especially those of us who are born blind, receive them
in the area of the tactual sense by the help of the groping
finger points. This olfactory sense of the insects, which
involves so remarkable a power of exact localisation in
space, has been described most adequately by Forel as
a topochemical sense of smell.
Forel, and afterwards Wasmann, have shown that

THE MUTUAL RELATIONS OF ENGRAMS 105

many phases in the biology of the ant are greatly eluci-
dated by an accurate knowledge of the peculiar nature
of this topochemical sense, and by the idea that the
impressions acquired through it are retained in the memory
of the animals or, as I should say, act engraphically.

But in all directions precise investigation goes to prove
that the successive engrams of smell are, in respect to
their mutual ecphoric action, non-reversible. Of course,
each series of engrams may also be acquired in reversed
sequence. We are able to learn the list, " panis, piscis,
crinis, finis " in that sequence, as also in the sequence
" finis, crinis, piscis, panis." The latter by frequent
repetition is engraved on my organism as an associated
succession of engrams. But we have done nothing beyond
acquiring a second succession of engrams which exists
independently. Like most ants, an ant of the species
Lasius follows the same track forwards and backwards.
On the strength of chemical stimuli frequently repeated
and acting in two definite sequences, two successions of
topochemical engrams are then gradually impressed on
it, but these impressions are almost as independent of
each other as if a different road had been chosen for the
return journey, which also exceptionally occurs.

Ebbinghaus by ingenious experiments succeeded in ob-
taining extremely interesting and important data on the
mode of connection — that is, on the mutual ecphoric
action — of successive engrams. The method devised by
him consists of the learning by heart of meaningless lines
of syllables until these can be faultlessly recited, the
forming of new combinations out of the previously learnt
lines, and learning by heart of these new combinations,
and finally a comparison of the number of repetitions
and of the time-periods required for the learning by
heart of the differently prepared combinations. By this
method Ebbinghaus discovered, first, that " at repeated
creation of lines of syllables, the separate members not
only associated with their immediate successors, but
that connections were also formed between each member
and several of its more remote successors. " The strength "

io6 THE MNEME

of the connection is a decreasing function of the time
or of the number of the intermediate members, which
separate the respective syllables in the original line.
The function is a maximum for those members succeeding
each other immediately. The intimate nature of the
function is unknown, only it decreases for growing dis-
tances of the members, at first very rapidly, and after-
wards very slowly." A second important result of the
experimental investigations by Ebbinghaus was expressed
as follows : — " In fact, in the learning of a line certain
mutual connections of the members move in two direc-
tions, backwards as well as forwards. The strength of
the disposition thus created was again a decreasing function
of the distance of the members in the original line. At
equal distances, however, it was considerably less for
the backward than for the forward connections. On
lines being repeated in equal frequency it was found that
each member is connected with the one preceding it
not much more closely than with the next hut one succeed-
ing it ; and with the next but one preceding it, hardly
as closely as with the next but two succeeding it."

Ebbinghaus has elsewhere calculated the ecphoric
influence of a member on the next but one succeeding
it at a third of its ecphoric influence on the immediately
succeeding member, as shown in the diagram : —

Preceding Engrams. < > Successive Engrams.

y ^ a b e

If we express the ecphoric strength in the successive
association of a on 6 by i, the ecphoric strength of a on
c is one-third, or expressed in the form of an equation : —

if a on 6=1

and a on c=\

and a on c=a on /3

then a on )5=J or a on 6=3 (« on )3)

according to Ebbinghaus.

One may therefore accept this calculation as also
representing the relative ecphoric influence of a member

THE MUTUAL RELATIONS OF ENGRAMS 107

on its successor and on its predecessor, and say that in
the case under observation, the ecphory of an engram
has an ecphoric effect on the engram succeeding it thrice
as strong as on the one preceding it.

Later investigations carried on by Miiller, Schumann,
and Pilzecker have confirmed in all essential points the
results obtained by Ebbinghaus on "nonsense-syllables."
Wohlgemuth's experiments gave identical results for
syllables learnt aloud. This author, however, inferred
from his experiments with optical stimuli that associa-
tions of visual engrams have equal ecphoric value in
both directions, backwards and forwards. Wohlgemuth's
extension of the experiments to visual engrams is un-
doubtedly of great value. They might with advantage
be made to embrace the remaining classes of engrams.
In view of Wohlgemuth's experiments, it seems to me
certain that there exist differences depending on the
nature of the sense-organ which serves as the receptor
of the stimulus and transforms it into the excitation
which acts engraphically. There are, however, equally
good grounds for assuming that these are only differences
of degree, and that they are in no way fundamental.
We have first of all this fact, which is i^ no way shaken
by Wohlgemuth's experiments, that we are absolutely
unable to reverse a long mnemic succession of stimuli —
whether visual, auditory, or tactile ; and secondly, we
find that if the stimuli which originated the series of
engrams in question are repeated in the original order
there is an immediate recognition of the succession as
famiUar ; but when they are repeated in the reverse
order, the succession strikes us as new and strange, even
though it consists of the characteristic movements of
some familiar object or the tones of some well-known
melody. Wohlgemuth's experiments on visual engrams,
confined to two consecutive members of a series and
neglecting longer courses, do not appear to me to affect
these fundamental facts. It would be well for us to
await the results of the continuation and extension of
these promising experiments, suggesting meanwhile that

io8 THE MNEME

they be carried on under a variation of the experimental
conditions, especially as regards time and the intervals
of exposure. I might be rightly expected to show where
we may look for the real cause of the unequal ecphoric
strength of the successive connection, for this is not
self-evident in the diagram given on page loi. Why
does the ecphory of E(syn) through ^^'(ac) not act as
strongly ecphorically on D(syn) as through ^'(ac) on
F(syn) ? I have dealt more fully with this problem in
the Mnemic Sensations (p. 205-216), and believe I
have found a satisfactory solution which, however, cannot
here be explained in detail.
The successively connected inherited dispositions also
"^ show an unequal ecphoric strength similar to that of the
individually acquired engram successions. It has already
been argued above that we have many reasons for treating
inherited dispositions as engrams also. Sufficient justi-
fication of this view will be given in the third part of
this book.

In any case, we find the series of all such inherited
dispositions, in respect of the mode and strength of their
connections, subject to the same rules as the succession
of the individually acquired engrams. A reversion in
the course of the series of reactions by which those dis-
positions manifest themselves does not occur, and for
them also the strength of the connection is a decreasing
function of the intermediate members, and is at its maxi-
mum for the members immediately succeeding each other.
The ecphoric action of member a on 6 is stronger than
on c, of 6 on c stronger than on d. It is therefore impos-
sible for a to overleap b and ecphorise c, or for b to
ecphorise d to the utter neglect of c.

A burrowing wasp (Sphex) burrows a hole, and then
flies out to catch insects, which, paralysed by the sting,
are dragged to the entrance. The wasp, however, before
taking its prey into the hole, always enters it first to see
whether things are all right. Whilst the wasp was in
the burrow, Fabre removed the insect to a short distance
away. When the wasp came out again, the prey was

THE MUTUAL RELATIONS OF ENGRAMS 109

found aiid again brought to the entrance of the burrow,
when the instinct to examine the burrow before depositing
the prey in it reasserted itself, and the insect was once
more left at the entrance. As often as Fabre removed
the prey, so often did the other action follow, so that
the unfortunate wasp had to examine the burrow forty
times.

Let us now analyse the case more precisely, denoting
the successive reactions of the dragging along of the
prey, by a, the laying down before the burrow, by b,
the examination of the burrow, by c, and the carrying
into the burrow, by d. The inherited dispositions corre-
sponding to these reactions let us denote by a, /3, y and 8.
We then have the following clearly ordinated succession
of engrams : —

Dragging along — Laying down — Examination — Carrying in.
a /3 y S

Fabre's experiment shows that when the wasp, after
manifestation of the series of dispositions a, /3, y, is placed
in a situation which renders a fresh manifestation of o
necessary, the manifestation of a commands that of /3
and that of the latter the manifestation of y. That is,
if we conceive the dispositions as engrams, the earlier
ecphorises the next. The association of disposition (or
engram) o with 8 is so much weaker than with j8 and y,
that, even with the experience of forty fruitless journeys,
the wasp was unable to effect the direct ecphory of'S
by a, and dispense with the ecphory of )3 and y. We do
not wish to convey the idea, however, that this succes-
sion is eternally immutable for Sphex and its descendants,
and that it could never be altered under the influence of
synchronically and engraphically acting stimuli. On the
contrary, in the Hymenoptera as well as in the Verte- ^
brates, we meet with many cases where inherited series
of engrams are liable to become transformed by new
individually acquired engrams.
We note exactly the same laws of association in the

no

THE MNEME

successions of those inherited dispositions whose reactions
manifest themselves as phenomena of growth. This fact
is of great significance for the understanding of the normal
ontogenesis of the organisms as well as of that experi-
mentally disturbed. As this subject will be dealt with
more in detail later on, I shall for the present leave it,
and in conclusion discuss one special peculiarity of suc-
cessive association.

A succession need not always continue in one series.
^ In many cases a line of excitations divides at a certain
point into two branches, and later, perhaps, into three
or more. A simple case is the distribution of the engrams
of a mnemically preserved piece of music that begins in
one part and develops into two parts.

Phase 1

2

8

4

8

C —

d —

'<

1
d

C —

1

e —

It is evident that in these equal phases the members of
the two branches of the bifurcation or dichotomy are
simultaneously associated. We may, therefore, describe
such a dichotomy of a series of engrams as a simultaneously
associated dichotomy.

But there also exist dichotomies, trichotomies, etc.,
of such successions, where the members of equal phases
are not simultaneously associated, and where the ecphory
proceeds at the point of division only in one direction.
This singularity often depends on the circumstance that
up to that moment no simultaneous association has taken
place. For instance, if I study separately the first and
second parts of the above-mentioned piece of music
without hearing, reading, or playing them together, the
dichotomy is not simultaneously associated, and therefore
liable, not to a simultaneous, but only to an alternating,
ecphory. The deliberate creation of simult9.p§ous asso-

THE MUTUAL RELATIONS OF ENGRAMS ill

ciations, however, changes the alternating into a simul-
taneously associated dichotomy. Numerous cases, how-
ever, exist, where such a transformation for one reason or
another is impossible. Generally, the difficulty follows on
the inability of the organism to carry out simultaneously
the reactions of both Hnes of excitations. In that case the
alternating dichotomy becomes permanent. On hearing
or reading, for instance, the end of the first stanza
of Fitzgerald's famous translation of the Rubdiydt of
Omar Khayyam now in the first and now in the second
version, I receive the impression of the following alter-
nating form : —

<in a Noose of Light,
with a Shaft of light.

Now, it is impossible to transform this alternating
dichotomy into a simultaneous associated one, either by
simultaneous hearing or by engraphic influences, such as
those of the above-mentioned two-part piece of music.
And wherever, for some reason or other, a simultaneous
association of the branches is impossible, the dichotomy,
or as the case may be, trichotomy, remains permanently
an alternating one. To the deeper understanding of the
nature of alternating dichotomy, acquaintance with the
conception of homophony is essential. Here I should
simply like to note that at the division-point of each
dichotomy, the complex of engrams borders on two or
more succeeding complexes instead of one. In the
above-quoted poem, for example, the word turret borders
on " in " as well as on " with." As this is clearly a case
of a dichotomy which only allows an alternate ecphory,
we are left to wonder whether the engram " turret " will
act ecphorically on " in " or on " with." Little doubt
may be indulged, if of the two associations one is closer
and, therefore, stronger ecphorically than the other.
But often merely a variation in frequency of the repetition
will give one branch of an alternative the dominance
over the other, Beside? thisj cQufljgting side-influences

112 THE MNEME

may operate in such a manner that the choice falls now
this way, now the other. We shall dwell on this mor6
fuUy in Chapters XII and XIII.

So far we have taken into account only the association
of engrams which present themselves as a legacy of ex-
citations produced by original stimuli. For the sake of
simplicity, we have written as if the simultaneous ex-
citation-complex, which at its disappearance leaves a
corresponding simultaneous engram-complex behind, con-
sisted simply of original excitations. But this is an arbi-
trary simplification. The simultaneous excitation-complex
usually contains, besides numerous original excitations of
all kinds, mnemic excitations ; and these as well as the
original excitations enter into the constitution of the
corresponding simultaneous engram-complex. Therefore,
in addition to the original excitations, all the mnemic
excitations ecphorised at the moment also belong to each
simultaneous excitation-complex and manifest the same
engraphic activity.

CHAPTER V

THE LOCALISATION OF EN GRAMS

So far we have made no attempt to deal with the nature
of the change which the irritable substance undergoes
between the emergence and subsidence of the excitation,
and which, with specific modification of the substance,
remains fixed in the engram. All we can say is that
the change in the irritable substance is most certainly
a material change. In limiting ourselves to a determina-
tion of the orderly connections between stimulus and
reaction, and in rejecting the lure of molecular mechanistic
hypotheses to penetrate into the " real nature " of the
excitation, we had no wish to rid ourselves of the obliga-
tion to enquire whether, within the same individual, the
irritable substance possesses in all its parts the same
properties ; but this question having long ago been
decided in the negative, we can now pass on to consider
the allied problem as to the way in which the irritable
substance, according to its heterogeneous properties,
distributes itself in the individual.

Long ago, and from widely differing starting-points,
the solution of this problem was attempted by physi-
ologists. Here the problem will occupy us only in so far
as it concerns the relation of the irritable substance to
the engraphic effect of stimulation ; that is, so far as it
touches the localisation of mnemic phenomena in the
individual. By approaching the problem of localisation
from the mnemic side, we shall be able to show it in a
special light, and at the same time deepen our insight
into the nature of mnemic phenomena.

8 «3

114 THE MNEME

Introductory Considerations on the Localisation
OF Inherited Engrams

A Planarian is a worm standing rather low in the scale
of animal evolution, yet possessing a differentiated central
nervous system — a brain and longitudinal nerves, two
eyes, a complicated intestine, and a genital system. If
one cuts the animal into sections at random, each section
from whatever part of the body is able, if it be not too
minute, to regenerate itself into a complete, worm, with
all its morphological and physiological peculiarities, and,
of course, with all its so-called instincts. Sections cut
from any portion of the body of Hydra, with the excep-
tion of the tentacles, regenerate the whole animal ; but
the size of the pieces must not be less than J to -^ mm.
in diameter, that is, about ^^ of an entire Hydra.
Segments cut from any part of the roots of many plants,
as, for instance, Scorzonera, Leontodon, etc., are capable
of building up the whole plant-individual, just as any
segment of a Begonia leaf, placed on moist soil, will
develop into a complete plant. Infusoria also, such as
Stentor, can be cut into many pieces, and if these are not
too small, and if they contain some portion of the nucleus,
a complete, though reduced, Stentor will grow out of each
piece. In adopting the view that the majority of the
inherited dispositions are engrams, we are justified in
thinking that the different sections of Planarian, Scorzo-
nera, Begonia, Stentor, possess in its entirety the inherited
engram-store of the complete individual. These segments
cut almost anywhere and anyhow are capable of a full
development, just as are the germ-products of the respec-
tive organisms.

In this connection it would be well to make more
explicit the idea that the germ-products and sections
cut from the bodies of certain animals and plants contain
the entire inherited engram-stock. This possession does
not imply the capacity to ecphorise the engrams at all
times and under all circumstances, for quite definite
external and internal conditions — or, as we would say.

THE LOCALISATION OF ENGRAMS 113

a specific energetic condition — are essential to the ecphory.
We are quite justified, therefore, in claiming that the germ-
ceUs, like the body-segments mentioned above, possess
also those engrams which are ecphorised only in the fully
developed animal or f)lant, where alone the potentialities
for it exist.

But from the fact that sections of certain organisms,
cut at random, are in possession of the entire inherited
engram-stock, we infer that in these forms, at least,
the engram inheritance is not localised in special areas
of the organism, but belongs everywhere to the irritable
substance of the organism.

There is a point in the division of the organism beyond
which further division becomes impossible without touch-
ing the vitaUty of the section, and thereby its capacity
for regeneration. It is obvious that a section which is
too minute for the absorption of nourishment from with-
out, and which is hardly able to answer the demands
made on it in the continuation of its life-processes, will
not be able to furnish the working capital for the purposes
of regeneration. The above-quoted experiments, however,
do not help us much in determining the actual minimum
size of those sections which, despite the mutilation, still
retain the entire inherited engram-stock.

But observations on the propagation of the organisms
prove that the entire inherited engram-stock may be
possessed by single cells, namely, the germ-cells. The
fact, however, that one can cut up indiscriminately an
infusorian like Stentor into various parts, each of which,
if it has but a small portion of the nucleus, will regenerate
an entire Stentor, argues strongly in favour of the view
that at least in this and similar cases the possession of
the entire inherited engram-stock belongs to a still smaller
biological unit than we get in the cell. With the problem
whether within each cell the engraphic change localises
itself predominantly or exclusively in the nuclear sub-
stance and its equivalents, and whether consequently
we have to conceive of the nuclear substances as being
the bearers of the engram — to which conclusion a good

ii6 THE MNEME

many data point— I shall not greatly concern myself,
as the problem is hardly ripe yet for discussion.

It may be well to state here the assured results of
our preceding observations. First, each germ-cell, or its
equivalent, which initiated each individuality phase,
possesses the entire inherited engram-stock. Most prob-
ably neither the cell nor even the nucleus of the cell is
the smallest unit able to possess it. For the sake of
brevity, we shall call the smallest unit able to contain
the entire inherited engram-stock the " mnemic pro-
tomer " ; whether it be the cell, or whether it be a more
minute morphological unit, we shall leave future research
to decide.

Secondly, we find that segments taken at random
from parts of a multicellular organism, whether plant
or animal, show themselves in numerous cases to be
possessed of the entire inherited engram-stock. But in
these cases also we are not yet in a position to determine
the morphological limits of the smallest mnemic units
or protomers, that is, the minimum segments of irritable
substance which still retain possession of this inherited
engram-stock. We cannot in these experiments depend
on extremely minute organic particles, because, compared
with the specially adapted germ-cells, they are placed
at a decided disadvantage for the double task of main-
taining the normal life-processes and of satisfying the
abnormal demands of regeneration. Consequently, the
question whether, under certain circumstances, a uni-
cellular segment of a fully developed multicellular organism
can regenerate the whole, and whether it therefore possesses
the entire inherited engram-stock still remains unanswered.
Evidence in the affirmative, however, can be adduced
in many cases from the phenomena of germ-development,
especially that of the segmentation of the germ-cells.

Further discussion of the problems connected with the
restriction of the regenerating capacity may well be deferred
until we are better acquainted with the nature of homo-
phony and the working of the developmental processes.

In Chapter XI of this book we shall again take up

THE LOCALISATION OF ENGRAMS tvj

the thread dropped here and shall have to concern
ourselves with the question whether the restriction of
the regenerating capacity, which is in a certain relation
with the processes of development, points to a definite
locaUsation of the inherited engram-stock or not.

Localisation of the Individually Acquired Engrams.

At the conclusion of the previous part we left it an
open question whether each cell or each protomer of
the growing, as well as of the fully developed, organism
possessed the entire inherited engram-stock. We begin
this section, however, with the statement that certainly
not every cell, not every mnemic protomer of the indi-
viduals quaUfied pre-eminently to acquire engrams, is in
full possession of the entire stock of engrams acquired
during the individual organic life.

We may affirm this with emphasis because we observe
that although the germ-cells are in full possession of
the engrams inherited by the organism from its ancestors,
only an exceedingly small number, if any, of the engrams
acquired in the individual life become manifest in the
next individuality-phase. Among the higher animals
neither father nor mother is able to transmit to the off-
spring in sufficient strength for manifestation the countless
engrams which during the individual life have been
acquired. Ability and power developed during individual
existence are not transmitted. We have seen, however,
in a previous chapter (p. 57-74) that the transmission
of individually acquired engrams from one generation
to another can in favourable cases be traced, especially
where impressions have been repeated through various
generations; but frequently the methods of observation
at our disposal render the proof of such a transmission
almost impossible. From the facts as we have them we
may argue to the probability of a transmission of indi-
vidually acquired engrams to the irritable substance of
the germ-cells, although with greatly weakened effect, and
through these to the progeny.

Xi8 THE MNEME

The engraphic action of a stimulus varies greatly in
relation to the different parts of the same organism.
Compare, for example, the engraphic action of single,
momentary, weak stimuli on the nervous substance of
the higher Vertebrata with the action of the same stimuU
on the germ-cells. The latter action seems so slight as
to be inappreciable. But, as we have already seen,
stimuli, in some cases, can be proved to act engraphically
on the irritable substance of the germ-cells. We are,
therefore, by no means justified in assuming straight
away from the negative results of certain experiments
the absence of all engraphic action.

A new light is thrown on the problem by the important
discovery of Tower (Evolution of Leptinotarsa, Washington,
1906) that the germ-cells of some insect organisms pass
through a period of extraordinarily increased suscepti-
biUty to stimuU. This sensitive quality, which is probably
not limited to the germ-cells of insects, but is far more
generally spread, accounts to a certain extent for the
capriciousness in the occurrence of hereditarily acquired
variations ; for the engraphic action of specific external
stimuli on the germ-cells during their period of increased
susceptibility would undoubtedly affect the variations.
(See The Problem of the Inheritance of Acquired
Characters, pp. 107-114.)

For the present, we leave the problem and state simply
this fundamental fact, that the irritable substance of the
germ-cells is in full possession of the inherited engram-
stock, and is thereby enabled to react with undiminished
force in the growing and fully developed organism. Of
the individually acquired engrams of the organism, how-
ever, the irritable substance of the germ-cells ' has but
few, and these of a vanishing order, so that hardly any
of the engrams acquired by the parental organism appear
in the later products, that is, in the growing and fully
developed organism.

Apart from the transmission of the greatly weakened
individual engrams to the irritable substance of the germ-
cells, are there any other data which tell in favour of

THE LOCALISATION OF ENGRAMS iig

a wide and varied distribution of engrams over the irrit-
able substance of the organism ? It is wise in deciding
this question to turn to those organisms which are especially
sensitive to engraphic stimuli, and which manifest the
change of state generated thereby by unmistakable re-
actions. First of these are Man and the higher Verte-
brata. In the second place come many of the Insects,
among whom some of the Hymenoptera, Bees, Wasps,
and Ants are the most important. Finally, there are
several Cephalopodes.

A strictly definite localisation of the individually
acquired engrams — ^the memory-images of the physio-
logists and the psychologists — ^has been accepted by a
great number of scientific workers, and has been located
for Man and the higher Vertebrata in the cerebral cortex.
The evidence for this Ues in the observation which has
been made over and over again that, in almost all the
pathological affections of the human Memory which
confront us in manifold forms, the only constant organic
alteration is a somewhat diffuse process of degeneration
in the cerebral cortex. In one case at least we are in
a position to define the region more precisely, so as to be
able to say with assurance that when a right-handed man
is suddenly deprived of his memory of words (amnesic
aphasia), a cerebral lesion has taken place where the
area of the left island of Reil borders the frontal and
temporal lobes. Formerly, many physiologists and patho-
logists went much further in the locahsation of memory-
images. The visual memory-images were, for instance,
in a definite place behind the fissure parieto-occipitalis.
The further development of regional demarcation meant
that, for the memory-images of landscapes, persons,
numbers, letter-signs, etc., there had to be a distinct
and separate area. Finally, each letter or number sign,
each single visual impression, was regarded as having its
own special cerebral area or cell. In similar fashion, the
centre for auditory memory-images, and especially for
the memory of the sound of words, was located in a
certain part of the first left temporal lobe. Throughout,

120 THE MNEME

one figured the cell as a kind of drawer in which the
memory of the sound of a definite word wias stored, a
naive conception entirely superseded among brain-special-
ists, but surviving in some cases in the wider circle of
biologists and medical men.

We seem, therefore, to be placed in the dilemma of
having either to reject altogether a localisation theory
which imagines that each single engram can be stored
up in a cerebral cell — or in a comparatively small complex
of cerebral cells — as in a separate drawer, or to admit
that in the human organism a special interdependence
exists between definite regions of the cerebral cortex
and the ecphory, or, as perhaps we ought to say, the pos-
sibility of the ecphory of distinct individually-acquired
engrams. The latter admission implies, however, the
recognition of a certain localisation, although it need not
be the kind which makes each nerve-cell of the brain a
repository for a specific engram.

In our preceding investigations we have repeatedly
emphasised the fact that of the innermost nature of the
process of excitation, as of the " nature " of any other
energetic process, we have no real knowledge. Our
hypotheses are dependent on similes for their intelligi-
bility. But while perforce we must be content with our
ignorance of the real nature of the excitation, we can
well concern ourselves with the principles that govern
its initiation, progress, cessation and after-effects. Here
at once we discover something very important for the
problems occupying us now, namely, the fact that the
process passes through the organism — ^most markedly
through those which are equipped with a well differen-
tiated nervous system — not in a vague, diffuse manner,
but along definite and fairly well isolated routes.

In this way the original synchronous excitation may
be regarded as already localised ; and as the engraphic
change of the irritable substance depends directly on
the s3mchronous excitation, a certain localisation of the
individually-acquired engrams within the irritable sub-
stance of an organism is implied from the beginning. If

THE LOCALISATION OF ENGRAMS 121

the isolation of the routes traversed by the excitation
were perfect, we should have a right to expect a perfectly
well defined localisation. Much experience, however, has
taught us that the isolation of these routes corresponds
simply to functional requirements, and is by no means
absolute. By " routes " I do not mean only the nerve-
fibres, but the whole tract through which the synchronous
excitation flows from its start until its cessation, whether
through nerve-cells, nerve-fibres, the grey matter of the
brain, or through other forms of irritable substance.
The whole tract of irritable substance covered by the
excitation may be described as the " primary area proper "
of a definite excitation. To the overflow of the excitation
beyond the usual boundaries of the area proper we may
trace the so-called reflex spasms, which can be liberated
by ordinary stimuli during a heightened irritability of
the central nervous system, such as is induced by strych-
nine-poisoning, tetanus, and hydrophobia, or by an
increase of the strength of stimulation during the normal
irritability of the nerves. These spasms have also been
called " disorderly reflexes " — a description not altogether
exact, as Pfliiger has shown that the overflow of the
excitation in the central organ does follow a certain order.
It manifests itself first in a contraction of those muscles
of which the motor nerves are on the same side of the
spinal cord, and at the same level as those which in
ordinary reflex action respond by contraction to the
stimulation of a definite place on the skin. With the
greater overflow, the nerve-complexes of the other side,
but only those which correspond with the affected nerves
of the primary side, are also affected, but in somewhat
less measure. Later, a stiU greater overflow means that
the nerve-complexes of the higher levels are influenced.

A relationship similar to that between the reflex spasms
and the ordinary reflexes exists between the automatic
movements and the so-called " co-movements." While,
for instance, it is easy to keep one arm quiet during
moderate movements of the other arm, control of the dis^
engaged arm during violent movements, such as fencing,

122 THE MNEME

has to be gradually learnt. But a man, whose irrita-
bility has been heightened by emotional influences or
by intoxication, will make such co-movements on the
slightest innervation, moving, for example, his left arm
during feeble movements of the right. Other evidence
can be adduced to show that the laws governing the
reflex spasms are valid also for the overflow of the excita-
tions in the case of automatic movements.

In the sensory area, the so-called irradiations of visual
and tactile excitations are also based on an overflow of
the excitation beyond its primary area proper. It tells
of a particularly imperfect isolation of the " Hnes," or,
as we may call them, the conduit, if, when the outer ear
passage near the tympanic membrane is touched, a tick-
ling is felt in the larynx; it is well to remember that
both regions are supplied with vagus fibres.

We see from these examples that the isolation of nervous
conductors is far from being absolute. A relative isola-
tion may be inferred when during normal irritability,
and with very weak stimuli, the overflow of the excitation
beyond the area proper does not become manifest. But
as soon as the stimulation exceeds a certain measure or
the irritability passes beyond the normal, the overflow of
the excitation becomes apparent. We may infer from this
that non-manifestation does not mean non-existence.

It is adherent in the nature of the process of excitation
that each excitation must first have attained a certain
strength, what may be called a threshold value, before
it becomes manifest to us by reactions. In our considera-
tion of the summation of stimuh, we saw (p. 33) that
an extremely weak stimulus might very well generate an
excitation without the latter necessarily manifesting itself
to us.

The action of an excitation appears to be confined,
as a rule, to a definite primary area proper, which varies
according to the stimulus. How this comes about is
entirely unknown to us. Especially are we ignorant of
the action of the excitation in regard to the peripheral
nerves and to the grey matter of the central nerve-organ.

THE LOCALISATION OF ENGRAMS 123

which is without the isolating structure of the white
matter. The action of the excitation is not, however,
hmited to the primary area proper. It affects the re-
mainder of the irritable substance of the organism, spread-
ing at first over adjoining, thence over more distant,
Unes of the irritable substance, until at last it permeates
the whole body. Observations on reflex spasms, on
certain co-movements, and on sensory radiations, con-
vince us that this overflow of the excitation beyond its
natural primary area proper, although it varies from case
to case, takes place in a quite definite order, the excitation
decreasing in strength proportionately to the distance of
its secondary sphere of action from the primary area
proper. In this manner the excitation finally spreads
throughout the whole irritable substance of the organism,
in gradually decreasing strength indeed, but still in
sufficient vigour to affect those parts of the irritable sub-
stance which are at the greatest distance from the area
proper of the respective excitation, or which are only
indirectly connected with it. Later, we shall explain
why we feel compelled to adopt this view.

We have now reached the point from which we can
obtain an insight into the phenomena of mnemic locaUsa-
tion. It may be well to restate what seem to us the two
fundamental suppositions from which it is allowable to
infer the existence of a certain mnemic locahsation within
the organism.

The first supposition concerns the engraphic action of
the excitation, that is, the specific change which remains
in the irritable substance after the lapse of the syn-
chronous stimulation. The engraphic effect is in a definite
relation to the strength of the synchronous excitation.
Very weak excitations seem to leave behind no engraphic
effect. But the case is really otherwise. For an engraphic
effect becomes manifest in the frequent repetition of weak
excitations, which proves that each factor must have
had an engraphic effect proportionate to its strength.
The second supposition was that within the more highly
developed organisms, especially Vertebrata with a specific-

124 THE MNEME

ally differentiated nervous sjretem, the original excitation
does not affect the whole irritable substance of the organ-
ism equally, but that it reaches the climax of its strength
in the specific section which we have described as the
primary area proper of the excitation. From there it
radiates with decreasing force, following the definite
paths into more and more distant areas of the irritable
substance, exercising a scarcely appreciable influence on
the most distant ones, and excluded entirely, perhaps,
only from those which by too specific a formation remain
altogether outside the sphere of its influence.

It follows from these suppositions that every excitation
taking place in the organism in sufficient strength will —
with the possible exception just mentioned — engraphically
influence each cell or mnemic protomer, as we have caUed
it, in a degree varying in strength only according to
the position of the protomer within the organism. For
example, a protomer located in those parts which com-
prise the area proper of the gustatory excitations will
be strongly influenced engraphically by gustatory stimuli,
but not at all strongly by other kinds.

When, therefore, at any given moment, not merely a
single stimulus, but as is normally the case, an entire
complex of numerous photic, auditory, tactile and other
stimuli acts upon the organism, this complex of simul-
taneous stimuli will, as such, influence every cell or
mnemic protomer ; but the influence will vary propor-
tionately to the relative positions of the protomers.

To make the point still more clear, it may be per-
missible to use a simile. But in referring specially to
phonographic reproductions, I should like it to be dis-
tinctly understood that I do not in the least intend to
suggest thereby any analogy whatever between the
genesis of an engraphic change of the organic substance
and the production of a phonogram. An organic engram
stands in much the same relation to a phonogram as
does a horse pulling a carriage to a locomotive propelling
one. The results under certain circumstances are similar ;
but the ways and means by which they are achieved are

THE LOCALISATION OF EN GRAMS 125

fundamentally different. But as it is perfectly legitimate
to compare the work done by an engine with that done
by a horse, perhaps I may be permitted, without fear
of misconception, to emphasise the idea of the topo-
graphical peculiarities of the action of complex influences
on the organic substance by referring to the topographical
peculiarities of the action of complex acoustic vibrations,
as illustrated in the making of certain phonographic
records.

Let us imagine that in an opera house of the usual
construction a great number of very similar phonographic
recording machines are distributed in different parts of
the building, among the boxes, the stalls, the dress and
upper circles, on and behind the stage, and also in the
orchestra between the seats of the players. In the
separate reproductions of the various records made
during the playing of the orchestra it will be found
that no two of the records are alike, despite the simi-
larity of the machines. According to the location of the
machines, it will be possible to distinguish differences of
clearness and power in the reproduction of the music.
Among the instruments distributed in the orchestra
itself, those in the vicinity of the basses will reproduce
the renderings of the bass parts out of all proportion
to the designed effect of the total production. The
phonographs placed between the 'cellos will in their
reproduction give us the impression that during the per-
formance the 'cellos played the leading part, and that
the rest of the instruments provided merely a pianissimo
accompaniment. So, with the records made by the other
machines, there would be differences of emphasis accord-
ing to their position.

The nature of mnemic reception and reproduction is,
of course, vastly different from that of phonographic
records, yet the results due to mere position and to the
related reception of complex influences are similar in
both the phonograph and the mnemic protomer. In
the latter case, however, we are concerned not simply
with the effect of acoustic influences, but also with photic.

126 THE MNEME

thermal, and electric influences, that is, with stimuli
belonging to all possible kinds of energies.

The strongest engrams which a mnemic protomer
receives are, naturally, from those excitations of a simul-
taneous complex in whose primary area proper it is
located. From the other simultaneous excitations of
the organism it receives more or less diminished vibra-
tions, which, of course, are stored engraphically in their
somewhat weakened character.

Again, within the area proper of an excitation, its
strength probably varies greatly according to its position
in the area. As yet definite data are not available. We
are quite justified in saying that the area proper of a
certain visual excitation extends over the retina, the
optic nerves, the chiasma, the external geniculate body
and thalamus, the upper corpora quadrigemina and the
connections with the eye-muscle nerves, and includes
certain cortical areas of the occipital lobe of the cerebrum.
Still, we have only very few data to decide whether quaU-
tative or quantitative differences of the respective excita-
tion-processes may be or must be recognised within this
area proper. We need not at this time attempt to prove
irom merely personal observations the probability that
within the area proper of an excitation there are local
differences of intensity. We shall begin with certain
definite assumptions and show that our observations find
an adequate explanation on the basis of these assumptions.

First, we shall assume that the strength of an excita-
tion within its area proper always reaches its maximum
in the irritable substance of the cerebral cortex ; and
we shall find that this assumption throws new light on
numerous groups of facts.

Again, by assuming that the cortex of the cerebrum
in the Vertebrata gradually evolved into a kind of " mul-
tiplicator " of the excitations, we can best understand
the exclusive position which, according to the findings
of comparative anatomy, of physiology, and of pathology,
it occupies in relation both to consciousness and to the
individually acquired engram-stock. For it is to those

THE LOCALISATION OF ENGRAMS 127

excitations which have entered into the cerebral cortex
and have reached their maximum strength there that,
from the point of view of introspection, conscious sensa-
tions correspond. And in regard to the individually
acquired engram-stock, we find that in the cerebral
cortex, where within their area proper the excitations
reach their maximum strength, there also are left the
most pronounced and most easily ecphorable engrams.
As, in accordance with the afferent system of its sensory
organs, the excitations of a sensory area reach the highest
degree of their development in a definite region of the
cerebral cortex, so the most precise engrams from this
sensory area will also be found in this region. Relative
to their distance from this region, the engrams show de-
creasing strength, while the engrams from other sensory
areas may increase in definiteness ; just as in the repro-
duction of the phonographic records derived from different
places in the orchestra we had the predominance of one
or the other instrument and the relative subservience of
the rest.

Our assumption adequately explains the correspondence
which, in the comparative study of the Vertebrata, we
find between the development of the cortex of the cerebrum
on the one hand, and the increase in stimulus-receptivity
and of engraphic power on the other. The perfecting
of the latter factors determines essentially what we are
accustomed to describe as increase in intelligence. Among
the warm-blooded animals, the individual acquisition of
engrams and their fuU use play a very important part.
The entire removal of the cortex of the cerebrum — ^this
multiplicator of excitations and principal retainer of even
fugitive impressions — Pleads to a notable damaging of
the individuahty, which is the more marked the higher
up the Scale of Being it occurs. Schrader experimented
with pigeons and falcons and Goltz with a dog. The
cerebrums of the creatures were removed. Their subse-
quent behaviour proves the truth of our statement.

From the same experiments we gather that in warm-
blooded animals, the individually acquired memory is

128 THE MNEME

predominantly, though by no means exclusively, localised
in the cerebrum. Although after the operation the
above animals no longer recognised either their own
kind, or their keepers, or their enemies, their hearing
and vision were practically unimpaired and they were
Still able to fly and to run. As the latter faculties are
by no means inherited, they must be considered as re-
actions, which prove the existence of individually acquired
engrams in places other than the cerebrum.

In the cases cited above the individually acquired
engrams, which on ecphory manifest themselves in the
complicated reactions of flying and running, are locaHsed
in subcortical portions of the central nervous system in
consequence of the frequent recurrence of these excita-
tions ; and their engraphic fixation is sufficiently strong
to allow of ecphory even in the absence of the cerebrum.
As to cerebral " localisation of symptoms " which has
been proved beyond doubt and has led to the great triumphs
of modern brain surgery, two possible explanations offer
themselves. Either an actual localisation of the engrams
themselves takes place, or the " localisation of symptoms "
is only a result of an interdependence between certain
definite parts of the cerebral cortex and the possibility of
the ecphory of specific groups of individually acquired en-
grams. Reference to this subject may be found on page I20.

Should future research demonstrate the fact of a specific
localisation of engrams, such a localisation could at the
most be but graduated, and in no case absolute, as has
so often been assumed. Such a graduated localisation
could be equally well explained by the above simile of
the phonographic machines. It may be that it does
actually exist to a certain degree, and that together
with a localisation of ecphory it furnishes the complex
phenomenon which confronts us in the cerebral " localisa-
tion of symptoms."

A more detailed explanation of these problems will be
given in a later consideration of the Mneme. At this
point I should like to show by quotation how C. von
Monokow in his recently pubhshed book on The Localisu'

THE LOCALISATION OF ENGRAMS 129

lion of Brain Functions faces this problem. " Before us
lies the difficult problem of determining the exact loca-
tion of the numberless stored up impressions, those
engram-complexes which manifest themselves as occasion
demands (mnestic processes). But the nearer we approach
the problem, the less sure we are that location can be
definitely determined " (p. 22). " The capacity to dis-
tinguish auditory impressions according to their nature
and intimate aesthetic significance, and the auditory
engrams built up thereon, have their proper working
areas in the entire cortex, although the parts lying in the
periphery of the so-called Heschl's convolutions (first
temporal convolution, sphere of hearing) are specially
important " (p. 25). " The working areas for the later
acquired engram-complexes (Semon), that is, what we
describe as 'perceptions,' 'presentations,' 'memory-
images,' etc., must, although differently distributed,
extend far beyond the proper somatic cortical areas over
the entire brain surface like a wide-spread fibrous tent "
(p. 27). "A certain local element, however, is essential
to all functions, even the highest, namely, that which serves
as physiological basis for the immediate realisation, or, as
Semon would say, the ecphory of various acts " (p. 27).

Whether excitations can be conducted from the body
surface centripetaUy to the central nervous system and
thence centrifugally over the entire irritable substance
of the individual, in strength sufficient to furnish in the
remote extremities recognisable engrams, is a question
exceedingly difficult to answer. In connection with this
problem the germ-cells will probably furnish the best
field for experiment and observation. The experimental
proof can be educed only by exposing the first generation
to the influence of a stimulus which reaches the germ-
cells only by the indirect way along which the excitation
is conducted. Should the generation developed from
these germ-cells then show itself changed in the above
direction without ever having been directly exposed to
the particular stimulus, it is evident that the engraphic
stimulus must have reached the germ-cells of the parent

9

130 THE MNEME

generation in a like indirect way, namely, along the path
on which the excitation generated by the stimulus was
conducted.

The conditions required are fulfilled by the experiments
already mentioned which Miss v. Chauvin made on the
Mexican newt, the Axolotl (Siredon), which is often bred
in Europe and kept in aquaria for amusement. These
Mexican newts are distinguished from their European
relations, the tritons and salamanders, by the fact that
at the end of their embryonic development they do not,
Uke other newts, lose their gills and make for land, but
under normal conditions they retain their gills and remain
in the water until they are sexually mature. They pro-
pagate, therefore, as fully equipped aquatic forms, that
is apparently as larvae. This is the rule. In Mexico,
however, natural influences similar to those experimental
conditions to which Miss v. Chauvin exposed her material,
seem to have acted on local varieties of the Axolotl ; for
these creatures, under natural conditions, breed varieties
with changed instincts.

The material, however, on which Miss v. Chauvin
experimented and to which we refer here, consisted of
a breed which in no stage of its natural development
showed any tendency to pass from gill to lung respira-
tion, and so to fit itself for land habitation. In animals
once sexually matured this transformation is impossible.

But Miss v. Chauvin was able, by appl3ang specific
stimuli at a certain critical stage of development, to force
the larvae to lung-respiration. The gills degenerated,
the larvae left the water, and finally passed by a perfect
metamorphosis into the gill-less land-newt (Amblystoma).
The method by which such changes of instinct and such
radical morphological alterations were effected consisted
of the enforced cessation of gill function and the stimu-
lated working of the lungs instead. This was somewhat
easily done by keeping the animals during the specific
time either in insufficiently aerated water, or by giving
them but a scanty supply of water. Contact with atmos-
pherie air puQices tp cojnpjete the metamorphosis.

THE LOCALISATION OF ENGRAMS 131

The newts coerced in this manner were kept alive until
as land-newts they matured sexually and propagated.
They deposited their eggs in the water, and the emerging
larvae passed through the ordinary stages of aquatic
development until they had reached the point where
the beginning of the metamorphosis becomes possible.
The ^nimals were then a length of 14 to 16 cm. To
obtain the metamorphosis of their parents, it was neces-
sary to place them under conditions unfavourable to
gill-respiration, but for the offspring this procedure was
not necessary. Although Miss v. Chauvin kept many
such larvae in well aerated water, they frequently came
to the surface to breathe and stayed there for hours —
behaviour which the Axolotl is wont to show only at
an advanced age and in water deficient in air.

Although further external coercion was withheld, the
subsequent course of the metamorphosis which Miss v.
Chauvin so induced in these animals was in kind essen-
tially different from and in speed far more rapid than
that which took place in the offspring of non-metamor-
phosed Axolotls of the same breed. Miss v. Chauvin,
therefore, felt justified in concluding " that this strongly
marked inchnation to continue development had been
transmitted to these individuals by inheritance."

To me such a conclusion admits of no question, but
I may also maintain that in this case there is an over-
whelming probability that the engraphic influence has
been transmitted to the germ-cells of the parent generation
by conduction, and that a direct stimulation of the germ-
cells by the respective stimuli is impossible. For the
germ-cells are embedded deep in the interior of the body,
where, unUke the cells of the external skin, they are not
exposed to changes due to the medium in which the
animal may Uve. Further, they are sheltered altogether
from the direct influence of all those stimuU, contact with
which is certainly involved during the transition from
aquatic to terrestrial Hfe.

It may be added that the germ-cells of terrestrial
Vert§br§ta are already surrounded by a moist medium.

132 THE MNEME

They lie in the abdominal cavity, and are thus always
flooded by the serous fluid. It seems to me, therefore,
altogether improbable that the osmotic condition of the
germ-cells should be affected by the fact that their bearer
lives as an Axolotl in water, or as an Amblystoma on
land ; for the latter, like all land-newts, invariably tries
to protect itself against too great a dryness of the medium.

Indisputable facts exclude the possibility that, in the
case of aquatic Amphybia, water might penetrate regu-
larly through the oviduct to the germ-cells. Kammerer,
who at first thought that this possibility was a reasonable
one, felt the force of my argument and abandoned his
original position.

Much more inconceivable is such direct physical in-
fluence in some of Kammerer 's experiments with Sala-
manders, especially where an hereditary influence on the
colouring was obtained by the action of light and humid-
ity, a stronger colouring of yellow in the offspring being
induced by keeping the parents on yellow earth. One
can quite understand that the comparatively small excess
of moisture to which the animal living on yellow earth
was exposed, in comparison with the one living on black
earth, might produce an effect on the skin immediately
exposed to the outside air ; but that this very small
difference in moisture should exercise through the bodily
tissues a determinating influence on the germ-cells em-
bedded in the permanently moist lymph space of the
abdominal cavity is altogether incredible.

In a still more definite way the argument applies to
the action of light, for the influence of light even on the
skin is indirect, requiring the mediation of the eye. By
blinding the animal in both eyes, it no longer reacts to
the different colour of its environment by change of
skin-colouring.

Many other cases of experimental research might be
cited to prove that the direct influence of a physical
stimulus penetrating to the germ-cells is impossible.

Nearly all animals kept in captivity gradually grow
tamer. Przibram, experimenting with the praying mantia

THE LOCALISATION OF ENGRAMS 133

(Mantis religiosa), noticed an increasing tameness in each
generation bred in captivity with complete exclusion of
selection. Where in this case is the physical stimulus
affecting the germ-cells directly ? In Chapter VIII of
my book, The Problem of the Inheritance of Acquired
Characters, readers will find a compilation of relevant
material.

It cannot be denied, however, that in many cases
physical and chemical stimuli may penetrate directly
-through the tissues of the body to the germ-cells. For
instance, in chiUing a plant or a cold-blooded animal,
the germ-cells are directly affected by the drop in tem-
perature. Chemical matter introduced into the tissues
of the body can directly affect the germ-cells, etc. Accord-
ing to Weismann and his adherents, such stimuU not
only act differently on the body proper, the " Soma,"
and on the germ-cells, but they also display correspond-
ing effects in both. Entering by specific receptors, and
making use everywhere of speciaHsed conductors, they
effect quite definite morphological and dynamic changes
in the Soma. StUl, they are supposed to be able to effect
in the germ-cells, quite independently and without the
intervention of such contrivances, a corresponding change
in the respective determinants of the germ-plasm. Detto
has applied the term " parallel induction " to these
hjTpothetical corresponding influences on the deter-
minants " of the germ-plasm, and on the soma with its
complicated contrivance for the reception and trans-
formation of the stimulus. But in spite of this sponsor-
ship he observes in the main a critical attitude towards
the hypothesis. Oiu: conception of an influencing of
the germ-cells through the soma by conduction Detto
calls " somatic induction."

The assumption a priori that all influencing of the
germ-plasm by the soma must be excluded led Weismann
to the theory of parallel induction. He asserts of the
germ-plasm that " its properties, chiefly its molecular
structure, do not depend on the individual in which it
is by chance embedded, for this, so to speak, is only the

134 THE MNEME

nutrient soil at the expense of which it is merely growing,
but that its structure is determined from the very begin-
ning." This conception of the physiological isolation of
the germ-plasm from the soma is not borne out by ana-
tomical facts, for histology has ascertained beyond a
doubt that there is a continuous organic connection
between the germ-cells and the soma, and that there
does not exist any isolating structure between the two.

As regards the physiological foundation of the theory
of parallel induction, I have examined it more minutely
in The Problem of the Inheritance of Acquired Characters,
and need not here do more than demonstrate, by means
of a single example, that as far as specific stimulations
are concerned, fundamental physiological difficulties
prevent our acceptance of the theory. Przibram and
Sumner independently discovered that the fur of rats
and mice, kept at unusually high temperatures, grows
thinner, while the peripheral organs, such as the ears,
feet, tail, external genital organs, increase in size. The
reverse takes place in animals kept at unusually low
temperatures. Under certain circumstances the changes
become hereditary, reappearing in offspring bred, bom,
and reared at mean temperatures. But how do these
experiments affect the physiological aspect of the problem ?

Variations of temperature, if not too extreme, act on
mammals principally through their skin ; the power of
the animal to generate heat tends to maintain the internal
organs at an equable temperature. Consequently nearly
all morphological reactions due to temperature variations
can be reduced to skin reactions. By continued ex-
posure to heat, the peripherally free parts like ears, tails,
hands, feet, etc., increase in size, but at the same time
a thinning of the hair takes place. These are in the
main specific reactions of the only organ directly affected
by heat, namely, the skin. The starting-point of these
changes is probably the extraordinary development of
the sweat glands and their excretory ducts ; and this
results in an increase of the superficies of the skin, and
the consequent partial shifting of the hair follicles with

THE LOCALISATION OF ENGRAMS 135

their sebaceous glands. Comparative anatomy teaches us
that the sweat glands are most developed on the soles
of the feet and on the palms of the hands, and that the
greater development of the sweat glands is in many
animals, including man, accompanied by an entire ab-
sence of hair. By continued exposure to heat these
glands increase remarkably in size. The tropical races of
many animals are destitute now of hair on their plantar
surfaces, but their near relatives in cold climates retain
the hair on those parts.

Under the influence of cold the contrary reactions are
noticeable, including not only diminution of the size of
sweat glands, but also an increase in the growth of
hair.

Apart from the more general effects on bodily size
and rate of development, we here meet on closer analysis
with a number of purely local and specific effects of stimuU.
And yet with these facts before us, we are asked to assume
that a problematical heating of the entire germ-cells by
a comparatively shght increase of the temperature, which
affects the germ-plasm directly without the mediation
of the locahsed and differentiated receptors of the skin,
has generated an exactly corresponding and exclusive
effect also on the skin of the offspring. This specificatibn
of the stimulation and especially its locaUsation on the
skin seems to me to dispose of the fallacious conception
of parallel induction.

On the other hand, the assumption of somatic induction
presents no such difficulties, for it simply regards the
individual with its soma and germ-cells as one organic
whole. The " Soma " furnishes for the whole organism,
including the germ-cells, indispensable contrivances for
the reception and transformation of the stimulus into
specific excitations, and this explains the homogeneousness
of their effect on the parental organism as well as on
the offspring. The only supposition is the sufiicient
susceptibihty of the irritable substance of the germ-
cells to respond to the excitations transmitted. Tower's
discovery of a period of increased susceptibility of the

136 THE MNEME

germ-cells promises to bring within closer reach the
solution of the many difficulties connected with an
hitherto capricious hereditary transmission.

Our investigations into the localisation within the
organism of the hereditarily-transmitted and the individu-
ally-acquired engrams may be summed up as follows : —
The data of regeneration and of experimental embryology
teach us that each cell, or rather each mnemic protomer
of a developing as well as of a fully developed organism,
is in possession of all those engrams which the organism
as a whole inherited from its ancestors. Of course, it
does not follow that every mnemic protomer is able at
all times to allow of the ecphory of these engrams, that
is, to reproduce always the corresponding state of excite-
ment, for this may reqtiire the presence of a quite definite
energetic condition.

The engraphic influences which affect the organism in
its individual life act, it is true, on each single protomer
of the body, but according to the mode in which the
stimulus enters the organism and to the changes which
the resulting excitation may undergo during its conduc-
tion, the engraphic influences act in varying strength in
the different protomers and according to their local dis-
tribution. The most marked differences in the locally
varying influences on the protomers of the same individual
occur in organisms with a highly differentiated nervous
system, where certain portions of that system may
become a kind of multipler of the excitations. The
protomers in these regions are the most strongly in-
fluenced engraphically. How the mnemic localisation
phenomena observed in Man and the higher Vertebrata
can be explained on this basis has been already dealt
with. Our conclusion may be now briefly stated. Each
protomer of the body receives most of the engrams
acquired by the organism during its individual life, but
according to the position of the protomer one group of
engrams is received more strongly, and another group
less strongly, than, by the protomers situated in other
parts of the body. Those protomers which in the higher

THE LOCALISATION OF ENGRAMS 137

Vertebrata appear to be chiefly favoured owing to their
position are situated in the cortex of the cerebrum.

By a less direct method we can assume from various
data of comparative anatomy and physiology that the
protomers of the cortical layer of the upper pharyngeal
gangUon in insects, especially in the Hymenoptera, and
the cortical layer of the cerebral ganglia of the Cephalo-
podes occupy a similarly favoured position. A single
excitation may suffice to produce a strong engram in
those protomers which, so to speak, lie in the respective
foci of these condensers. The irritable substance of the
germ-cells is situated altogether away from these foci.
The nervous excitations reach the germ-cells by many
roundabout ways, and generally very greatly enfeebled.
Frequent repetition in the individual life and through
successive generations raises these originally subliminal
engraphic effects above the threshold, that is, they become
hereditary engrams capable of manifestation.

In my opinion the numerous data of brain-physiology
and brain-pathology, as well as the data bearing on the
inheritance of acquired characters, may be interpreted
in accord with the view here set out on the localisation
of individually acquired engrams and on the mode of
their transmission to the germ-cells. To analyse in detail
from this point of view the enormous accumulation of
data concerning cerebral localisation and the phenomena
of inheritance would unduly extend the Hmits of the
present volume. But as I think it possible to carry out
this task on the basis already estabhshed, I hope to deal
with the subject more exhaustively in a subsequent work.

CHAPTER VI

ECPHORY OF THE ENGRAM. THE TWO
PRINCIPAL MNEMIC LAWS

By the ecphory of an engram we understand the passage
of an engram from a latent to a manifest state, or as
we might say, the rousing into action of a disposition
created by the original excitation and characterised as
a permanent but usually latent change in the organism.

The excitation resulting from the ecphory of an engram
we term " Mnemic excitation," and we have no reason
to assume that in its nature it differs essentially from
its predecessor, the original excitation. The differences lie
in the originating conditions.

The original excitation-complex is generated and main-
tained by the action of a stimulus-complex synchronous
with the excitations, and described by us as " the original
stimulus-complex." The corresponding mnemic excita-
tion-complex may be released by the partial recurrence
of this stimulus-complex. The ecphoric factor, therefore,
consists of the partial or entire repetition of that energetic
condition which formerly acted engraphically. The final
exposition of the process will be given towards the end
of this chapter. At this point it may be well to elucidate
the differences between the production of an original
and the ecphory of a mnemic excitation by the help of
the following diagrams, of which the first illustrates an
original, and the second a mnemic, process in the three

phases, e, f, and g.

138

ECPHORY OF THE ENGRAM 139

Phaie'e x, I7ta%e f TTiase g

>» ' Phase e

PAase f

PJia&e g

^f-

StuuMimeattI

I
I

fHmn)

»/■(!,

->fS(m.n.)

1 >g*Cmn)

— I — >^i(mn}
-I >y6^faij

140 THE MNEME

In the original process the stimulus-complex ?'"'
generates in phase e the original excitation-complex
fi'"' (or), in phase / the stimuH a^~' generate the original
excitations /'"' (or), in phase g the stimuh t'""^ generate
the excitations g^~^ (or). Let us assume that of the three
successive excitation-complexes — e'"'' (or), f^~' (or), g'"'' (or)
the components with the same indices — e^, /', g', or e',
p, g', or e^, p, g^, etc. — are more intimately related with
each other than are those with different indices, as for
instance, e' with/' or/* with g^. As we saw on page 95,
the engrams generated by the components of like index
will form a more coherent association than will those
of different index ; that is to say, the successive ecphory
of one by the other will be stronger in those of like index.
The diagram of the original process suggests this by the
horizontally curved lines, whilst the genesis of the simul-
taneous association is expressed by the vertically straight
lines. Objection might be made to the use of these
connecting lines in reference to the original excitations,
on the ground that the existence of such connections
had so far been proved by us only for the engrams gener-
ated by these excitations. The engrams, however, are
simply and solely the products of the original excitations,
and for this reason we may refer the peculiarities of their
connections to the corresponding properties of the gener-
ating original excitations. The validity of this conclu-
sion may be directly established by observation. The
introspective method shows most clearly the intimate
connection of the original successive components in the
identical manner in which this connection appears later
in the ecphory of the corresponding engrams. For
example, when listening to a piece of music, we perceive
the closer connection of the successive excitations due
to the operation of the single melodic parts as compared
with the succession of harmonies by an immediate process
in consciousness ; just as at an operatic performance
we note the closer connection of the words or of the music
as compared with words and music together, A further
illustration is afforded by the ballet. At the sight of

ECPHORY OF THE ENGRAM

141

two figures dancing simultaneously but separately, the
characteristic movements of each dancer are more strongly
related in consciousness than are the combined movements
of the two.

With the conclusion of each simultaneous phase, the
respective excitations seem also at an end, but so far
as they have acted engraphically, they have simply
entered into a state of latency. When aU three phases
have run out, then on the presupposition of a strongly
engraphic action of all components three engram-com-
plexes remain behind in the following order and con-
nection : —

Simultaneous

Engram Complex

e

Simultaneous
Engram Complex

Simultaneous
Engram Complex

g

fi' (engr)
I

e3
I

I

I

»6

— /' (engr) —
I

I

I
I

I
-P

— ^'(engr)
I

I

_^3
I

I
-g'
I

I
-g'

These three successively associated engram-complexes
can be ecphorised from the simultaneous complex e by
the appearance either of the original excitations e'-7(or),
or of a single component of this complex. In our second
diagram on page 139 we selected the single component
e* (or), to be generated by the stimulus component s*.
It might be asked why I do not simply say that the ecphory
results from stimulus s*, but I prefer the expression used
for the following reason : As the successive ecphory
ghows (see phases / and f in the second diagram on

142 THE MNEME

page 139), neither the stimuli, nor a fraction of them, but
only the respective corresponding excitations are neces-
sary for the rousing of a mnemic excitation. At the
ecphory, as was illustrated in the second diagram of
the mnemic process in phase e, we can, after the entry
of the stimulus ?♦, not only trace the mnemic excitation
c* (mn), but also the preceding original excitation e* (or).
This will be set out more fully in the next chapter, when
we consider the mnemic homophony. In view of the
fact that the engrams e', e'. e', e*, /""* and g*"* are ec-
phorised not by the original stimuli, but by the excita-
tions generated by the stimuU, we are fully justified in
assuming that engram e* (engr) also is not ecphorised
directly by the stimulus s*, but by the original excitation
c* (or) generated by s*.

The course of the mnemic process as indicated in our
second diagram on page 139 runs as follows : — The stimulus
s* generates the original excitation e* (or), which acts
ecphorically on the engram e* (mn). But the appearance
of the excitation e* (or) -f- e* (mn) involves the further
ecphory of all or of a part of the engrams simultaneously
associated with engram e*. We assume in our diagram
that it acts in sufi&cient strength only on a part, viz.
on the engrams e", e^, e*, and e*, but not on engrams e'
and e'. The ecphory of the sections /*"* and g*"* follows
by successive association on the ecphory of the sections
«'"*, in the manner suggested in the diagram. In the
course of the mnemic process of the case given, the
generation of the original excitation e* by the stimulus
s* produced the same or almost the same effect as the
joint action of the stimuli ?*"*, a*-*, t*"*, in the course
of the original process.

Next we have to consider a peculiar feature of the
ecphory which we have already suggested in our diagram
of the mnemic process (p. 139). The recurrence of
but a part of that energetic condition which acted en-
graphically may ecphorise the whole, or at any rate the
larger part, of the corresponding simultaneous engram-
complex ; but it need not necessarily do go, In fact.

ECPHORY OF THE ENGRAM 143

it nearly always ecphorises parts only, and although
these may far surpass the ecphorising excitation so
far as the number of components is concerned, they
hardly ever embrace all the components of the engram-
complex. So, referring to the diagram, we may say that
the original excitation e* (or) generated by s*. ecphorises
the mnemic excitations e'"* (mn), but is not able to
ecphorise at the same time the engrams e'(engr) and
e'(engr).

The reader may refer to the case cited in Chapter IV
(page 92). A definite olfactory excitation liberated by
the odour of boihng olive oil never fails to ecphorise in
me the associated engram " Capri seen from a certain
spot at Naples." This excitation, however, is not capable
of ecphorising other parts of the simultaneous complex,
as, for instance, the engram of the barrel-organ tune.
The latter ecphory occurred only when the identical
tune casually struck my ear again. To speak in terms
of our diagram, where the engram of the introductory
chord of the barrel-organ melody might be expressed by
e' (engr), the ecphory of this engram is possible only by
the recurrence of the original excitation e' (or).

We can easily convince ourselves that what we call
in ourselves and in our fellow-beings good memory is
only partially based on the facility and certainty with
which the stimuli act engraphically upon the organism.

For almost as important as the possession of numerous
and well-estabhshed engrams is their ready ecphory by
way of simultaneous and successive association in all its
manifold varieties and combinations. This readiness
implies a far more frequent rousing of the engram than
an ecphory dependent on the repetition of the original
stimulus which formerly had served for the generation
of the engram. An instance of the latter case is the
playing again of the forgotten melody, or the prompting
of the elusive word, which, for some reason or other,
" the weak memory " cannot recall. If the melody or
word is not then recognised, we may suspect that it has
never been fixed epgfaphically, Jhis supppsitipn is,

144 THE MNEME

however, not always beyond question, for in cases where
the energetic condition has greatly changed — ^as, for
example, in states of excitement, or during intoxication,
or in temporary amnesia — not even the recurrence of
the original stimulus suffices for the ecphory of the corre-
sponding engram. In other words, the excited, the in-
toxicated, or the insane man is not capable of recognising
objects well known to him at other times. Very in-
structive in this respect is the so-called periodic amnesia.
The cases of two independent memories, or as it is called
" double personality," are especially important to us.

I cite here a famous instance by Macnish {Philosophy
of Sleep) as quoted by Ribot, which has been confirmed
by later observations.

A young American woman lost, after unusually long and pro-
found sleep, the memory of all that she had learnt. She could
no longer spell, nor write, nor calcvilate, nor recognise the objects
and persons of her environment. A few months later she again
fell into a long and deep sleep. When she awoke she was once
more in the full possession of her previous knowledge and of the
vivid memories of her earlier life ; but of the intervening experi-
ence she had no recollection at all. During more than four yeaxs
she alternately passed from one state into the other, the passage
being marked by long and sound sleep. Of her double personaUty
she had not the slightest consciousness. In the one state she
possessed all her original knowledge, in the other only what she
acquired during her illness ; in the first state her handwriting
was firm and even, but in the second she wrote badly and clumsily,
as if altogether unaccustomed to the efiort. It sufficed not that
persons were introduced to her in either of the two states ; to know
them sufficiently, she had to make their acquaintance in both states.
And the same held good for all other things.

Similar states occur in cases of hysteria. The pheno-
menon of "alternating memory" may be observed in
hypnotised persons, who remember their hypnotic experi-
ence only while in the hj^pnotic state. In the waking state
the hypnotic interval is a blank. Alcoholic intoxication
may, under certain circumstances, create an energetic
condition whose engrams are ecphorable in the next

state of intoxication, but not in the intervening state

ECPHORY OF THE ENGRAM 145

of sobriety. A case cited by Ribot is instructive. An
Irish porter, who when drunk had lost a parcel, and when
sober could not remember where he had left it, recollected
the necessary facts during his next drunken bout. In
all these cases during the state of sleep, of hypnotisation,
of fever, or of mental disturbance — which we may call
state b — an energetic condition has been created, so
much differing from the state of waking, of soberness,
or of absence of disturbance — which we may call state
a — ^that the ecphory in state b of the engrams acquired
in state a is more or less defective. Only in cases where
by virtue of the experience of years the engrams are deeply
fixed and frequently ecphorised may we expect ecphory
independent of abnormal or contrasting conditions.

Similar conclusions are drawn by Ribot, who referring
to periodic amnesia (op. cit., p. 71) says that the "two
physiological states in their alternation involve specific
feehngs which determine two forms of associations, and
consequently two memories."

It is now possible to summarise in a graduated series
the influences which produce the simultaneous ecphory
of a definite engram. These are : —

1. The complete recurrence of the energetic condition
which ruled at the generation of the engram. This has
the strongest ecphoric effect.

2. The partial recurrence of the energetic condition
inclusive of the original excitation which generated the
engram. This partial recurrence, however, may be alto-
gether ineffective under certain circumstances, as when
the energetic condition has been greatly changed, by
intoxication, hypnosis, etc.

3. The partial recurrence of the energetic condition
exclusive of the original excitation which generated the
engram. In this case the remaining simultaneously asso-
ciated engrams act ecphorically with varying strength.

The two kinds of ecphory which we noted in our pro-
visional survey, and which we described as chronogeneous
and phasogeneous ecphory, are at bottom but specific
cases for which our general definition of the ecphoric

10

146 THE MNEME

influence is valid without qualification or addition. In
both instances the entire or partial recurrence of a definite
internal energetic condition acts ecphorically.

Chronogeneous ecphory takes place, as already noted,
in respect of individually acquired, as well as of inherited
engrams. Concerning individually acquired engrams, the
reader is reminded of the cases analysed on page 49.
If we are accustomed to take our first meal daily at eight
o'clock, and our second at one o'clock, we have usually
no feehng of hunger in the interval. But if, for some
reason or other, we get into the way of having a meal at
eleven o'clock, an insistent appetite asserts itself regularly
and punctually at that time, even when we are quite
occupied with other things. So also in regard to our
desire for sleep. A little nap taken in the midst of our
daily work soon establishes itself as a necessity. At the
usual time sleepiness overcomes us as a result of a purely
chronogeneous ecphory.

An interesting case of chronogeneous ecphory in the
Axolotl is reported by Miss v. Chauvin (op. cit., page 382)
For a period of three years and two months the animals
were kept during the day on land, and during the night
in water. Their development from gill to land newt was
thereby delayed, and an intermediate stage maintained.

To quote Miss v. Chauvin's words : —

One of the animals, for the purpose of its metamorphosis into
a land animal, had for some time to give up the water entirely.
Only when the exuviation had been effected was it permitted again
to make for the water. With the restoration of its liberty, the
animal, much to my surprise, went each evening into the water,
and each morning on to the land, at those times which previously
it had been forced to observe. This behaviour, very striking for
an Amblystoma, was kept up without interruption from the 20th of
January to the 25th of April. From that date onwards the animal
remained concealed in the moss at night also, and only returned
into the water when the desire for greater moisture had been
generated in it by the process of shedding its skin.

As an instance of the pure chronogeneous ecphory of
an inherited engram, we gave the unfolding and closing

ECPHORY OF THE ENGRAM 147

of the leaves of Mimosa and Acacia while they were
subjected for twelve hours to an even, continuous ex-
posure to light. The winter rest of beeches, and the daily
and yearly periods of many plants are examples of pure
chronogeneous ecphory.

The partial recurrence of the internal energetic con-
dition, we explained in all these cases by the cycle of a
definite association of metabolic processes, on the tempo
of which the chronometric capacity of the organisms is
founded. This chronometer, as we may call it, is of
course just as little infalhble as a manufactured one, and
will gain or lose according as the tempo of the respective
metaboHc processes is changed by external influences.
Under certain circumstances, however, it is very difiicult
to interfere with this tempo, as the behaviour of the
beeches during their winter rest proves. Here, the
chronometer is only slightly accelerated by the permanent
increase of temperature, although we might reasonably
imagine that such increase would very much accelerate
the speed of the metabolic process. But probably these
winter metaboUsms are to a great degree independent of
the external temperature.

Phasogeneous ecphory also is, after aU, nothing else
but the recurrence of a definite internal energetic con-
dition, and numerous facts given in the statistics of
development and confirmed by experimental embryology
teach us that frequently the partial recurrence of this
condition is sufficient. The reader is referred to what
already has been stated in my exposition of the factors
which Uberate the lens formation in embryonic develop-
ment. Even such great alterations of the internal
energetic condition, as are involved, for example, in the
experiments by Roux on the eggs of frogs, do not hinder
in the untouched half of the developing egg the appear-
ance of phasogeneous ecphory and the reactions of growth
which thereupon follow. As it is our purpose to deal
fully with phasogeneous ecphory and its peculiarities in
the following part, it is not necessary here to enter
further into the matter.

148 THE MNEME

The main results which we have reached in this and
the two preceding chapters may now be embodied in
the following laws, which between them contain the
quintessence of mnemic " first principles." They are the
two principal mnemic laws.

First principal mnemic law : Law of Engraphy. All
simultaneous excitations within an organism form a
coherent simultaneous excitation-complex which acts
engraphically ; that is, it leaves behind it a connected
engram-complex, constituting a coherent unity.

Second principal mnemic law : Law of Ecphory. The
partial recurrence of the energetic condition, which had
previously acted engraphically, acts ecphorically on a
simultaneous engram-complex. Or, more precisely de-
scribed : the partial recurrence of the excitation-complex,
which left behind it a simultaneous engram-complex,
acts ecphorically on the latter, whether the recurrence be
in the form of original or mnemic excitations.

As already indicated in the course of the last three
chapters, the laws of association can be deduced from
the above two laws in a simple manner. We therefore
add the explanatory proposition : " Association depends
on the conjunction of single engrams ; it makes its appear-
ance during their relatively isolated ecphory, and originates
simply from the presence of the respective components
in the same simultaneous complex." Association, there-
fore, is always simultaneous association.

CHAPTER VII

MNEMIC EXCITATION AND HOMOPHONY

In Chapter I we concluded that each original excita-
tion in its emergence, duration, and subsidence depends
on the emergence, duration, and subsidence of an
elementary-energetic condition regarded as stimulus.

Mnemic excitation is independent of such an elemen-
tary-energetic condition. The necessary conditions for
the emergence of a mnemic excitation are : first, the
existence of an engram whose nature determines the
nature and duration of the mnemic excitation ; and
secondly, the action of an ecphoric influence.

Each mnemic excitation is, therefore, related to an
original one, standing to it much as the reproduction
of a picture does to the original. In most cases, however,
the mnemic reproduction renders only the strongest lights
and shadows, for if the mnemic state of excitation is to
equal the original one in vividness, or perhaps surpass
it, an especially favourable state of the individual at the
time of ecphory is required.

The mnemic state of excitation reproduces the original
excitation in all its proper proportions, inclusive of time
values. This may adequately be gathered from the
study of mnemic successions, which, if no extraneous
disturbing element interferes, take place in exactly the
same rhythm as the succession of the original excitations.
In respect of their intensities and of the tempi of their
courses, the mnemic excitations reproduce or represent
the original excitations only in their proportional and not

J49

150 THE MNEME

in their absolute values. We shall treat this more fully
in Chapter XIV, where we discuss the proportional
mutability of the mnemic excitations.

Whilst the duration of the main phase of an original
excitation corresponds exactly to that of the generating
stimulus, the duration of a mnemic excitation is deter-
mined not by the duration of a simultaneous stimulus,
but by that of a previous stimulus, that is, of the one
that acted engraphically. The temporal limitation of
the mnemic excitation is therefore, in a sense, pre-
determined. Much the same qualification applies to
the intensity, but not to the vividness of the mnemic
excitations.

The mnemic excitation need not necessarily appear as
a link in a succession, for when released by simultaneous
ecphory, it may also appear as an isolated state, and
disappear without having influenced, at least in any
manifest way, an ecphoric succession. Readers may be
reminded of the image of a landscape, which, on the
perception of the smell of oil, arose and quickly faded,
but without necessarily having liberated in a perceptible
manner the series of successive and related mnemic
excitations.

But in this case, also, it may be maintained that the
mnemic excitation stands in a certain relationship of
temporal duration to the original excitation, for the
succession of mnemic excitations is really more or less
a matter of chance determinable by various factors ;
for example, the later engrams may be much less sharply
outlined than their ecphorised predecessor, or the course
of the ecphory of the mnemic succession may be disturbed
either by new original stimuli or by new associations.

It is of little use, therefore, tr5dng to trace the ex-
tended connections of individually acquired engrams, for
usually we are distracted by new synchronic impressions
and deflected into associated side-tracks. But certain
modes of auditory stimuU, the successions of melodic
tones or of Unes of poetry, produce in us such closely
conjoined engraphic connections that, on ecphory, they

MNEMIC EXCITATION AND HOMOPHONY 151

run throughout an unhindered course more frequently than
in any other series of engrams.

Possible disturbances and deflections may therefore
cut a mnemic excitation short as compared with the
original excitation. But if the mnemic excitations £ire
merely retarded, the retardation takes place with such
proportional exactitude for each member in the succession-
that the rhythm of the original excitations is always
maintained.

Mnemic excitation cannot be retained for any length
of time as an uninterrupted excitation. Where that
seems to be so, it is in reaUty a case of the repeated
ecphory of the one engram. A painter, for instance,
may work for hours on a portrait, the original of which
is no longer before him. What' is perceived by him is
an ecphorised engram, a " memory-image." And this
mnemic model is not the result of a single, but rather
of an oft-repeated, ecphory which later, as the painting
more truly resembles the subject, will be exercised by the
portrait itself.

In this connection we often meet with a combination
which affects the mnemic phenomena in a pecuHar manner,
and which demands careful analysis. The starting of a
mnemic process invariably requires an ecphoric impulse.
According to our second mnemic law, this impulse may
be the recurrence of either original or mnemic excitations.
The reader is referred to the diagram on page 139. Here,
the mnemic process follows on the recurrence of an original
excitation. In the diagram this excitation e*(or), which
is generated by the stimulus s*, first ecphorises the
mnemic excitation e*(mn), and this is followed by the
ecphory of the mnemic factors of the successive phases.
As the diagram shows, in phase e the original e'*(or)
and the mnemic excitation e*(mn) co-exist. But in the
later phases f, g, . . . the corresponding original excita-
tions are absent. An illustration may be useful. If the
reader hears the first phrase of a proverb, for example,

" A bird in the hand " he wiU mentally complete it.

The original excitations generated by the opening words

152 THE MNEME

act as stimuli, and ecphorise the corresponding mnemic
excitations which co-exist with them. But during the
mental reproduction of the remainder of the proverb
mnemic excitations only are involved, the corresponding
original excitations being entirely absent.

At present, however, we are less concerned with the
purely mnemic course from the sixth word onward than
with the simultaneous existence of the original excitations
generated by the five stimulating words, and of the mnemic
excitations ecphorised by these original excitations e*(or)
and e*(mn) in phase e of the diagram.

The co-existence of original and mnemic excitations
may be admitted, but this admission, as some one may
point out to us, does not imply that in their co-operation
each maintains undisturbed its independent action. For
their action may be a blend resulting in uniformity, or,
if slightly different, the one may affect the other, by
interference either weakening or strengthening it, but in
the conjoint effect showing uniformity, or, lastly, they
together may have only the effect of either by itself.

During our investigation of the original S5mchronous
and engraphic stimulation (pp. 90-98) we recognised as
characteristic of the synchronous action of stimuli on the
organic substance, that the various excitations resulting
therefrom do not lose their identity in a diffuse mixture,
but co-exist and run their course in juxtaposition. The
same may be said of the simultaneous course of a mnemic
excitation and of a new original excitation related thereto.

This , can be strikingly proved by reference to those
numerous cases where the new original stimulus, which
acts simultaneously with the mnemic excitation, is similar
to, but not identical with, the pre-vious original stimulus,
which, in its engraphic action, prepared the ground for
the mnemic excitation. The results of the incongruity
of the mnemic and the new original excitation can be
distinctly recognised by us, even in the most trifling
details. If after some years we visit a well-loved and
familiar scene, we are keenly sensitive even to the slight
changes that in the meantime may have taken place.

MNEMIC EXCITATION AND HOMOPHONY 153

We note the disappearance of this or that tree, the
presence of a new house, or the alteration of an old one.
A capable conductor, who knows well the work in hand,
and so is able to dispense with the score, notices with
astonishing minuteness the omission of this or the pre-
mature entry of that part or any slight variant that may
be introduced by the soloist. The incongruousness of the
mnemic with the simultaneously occurring original process
is immediately perceived.

Cases from other sensory areas illustrating the process
which ensues when mnemic excitation and the new
original excitation are brought, so to speak, into relations
of superposition, and where each incongruity causes a
reaction of sensitivity, might readily be given, if occasion
demanded. Our ordinary daily experience furnishes in-
stances in abundance.

Assuming, however, that the mnemic excitation and
the new original excitation coincide so closely that no
reaction of perception of difference follows, we may
prove by reference to the reaction of recognition that
both kinds of excitation co-exist, and that in running
their courses they do not blend into a unified whole.
In those cases where the mnemic and original excitations
are in perfect congruity, the reaction which follows is
pure in character ; but in cases of imperfect congruity,
the reaction is marked by the perception of difference.
In order that the reaction of recognition or of perception
of difference may arise as a clear process in consciousness,
the mnemic excitation must possess a certain vividness.
We may distinguish well-marked differences of state
when we compare the initiation of an original excitation
with the vague sensation of seeing, hearing, or feeling
something over again, accompanied by weak, mnemic
excitation, or when we compare either or both of these
with precise recognition and strong mnemic co-excitation.
Between these different states there are innumerable
gradations. For the fundamental problems under in-
vestigation, however, it is relatively unimportant whether
the reaction of recognition is strong or weak,

154 THE MNEME

By introspection, the simultaneous course in juxta-
position of the mnemic and the new original excitation
may be known in two very characteristic reactions, the
reaction of recognition, and the reaction of the percep-
tion of difference. This phenomenon of the simultaneous
course of the mnemic and the new original excitations
in juxtaposition plays a most important part in the
biology of the organism. It is useful, therefore, to call
it by a special name. After long consideration the term
" Homophony " has been chosen to signify the state of
unison in which (i) a mnemic and a new original excita-
tion, or (2) two mnemic excitations, or (3) two original
excitations may find themselves.

The term, of course, is literally accurate only when
the excitations belong to the auditory area. To the
other excitations — visual, tactile, olfactory, etc. — it can
only be applied in a metaphorical sense. But a metaphor
which gives wings to our imagination and does not pretend
to be an exact scientific description may, in the absence
of a term strictly applicable to the facts, fully justify
itself in the deepening of a scientific understanding of
the subject-matter.

It is, of course, far more difficult to discern the process
of Homophony from objective reactions in creatures
other than ourselves than from introspection, in which
we are aided by the reactions of recognition or of per-
ception of difference. In the case of our fellow-men, the
reactions may be made known to us by the agency of
speech, and language may be made sufficiently unequivocal
to indicate the existence of the mnemic homophony.

But the matter assumes a different aspect when we
have to deal with those organisms which cannot express
themselves in articulate speech, and which, in spite of
their intimate relationship with us so far as structure and
vital processes are concerned, are still too far removed
from us in other respects to warrant us in invariably
interpreting our observations on them in terms of our
own sensations. We have consequently to guard against
an injudicious reading of our states of consciousness into

MNEMIC EXCITATION AND HOMOPHONY 155

those of other creatures. In principle, I am adverse to
distinctions which set the Genus Homo over against all
other organisms, but, in the present state of our know-
ledge and means of observation, the only safe method
in the physiology of stimulus is to consider such reactions
as are accessible to direct observation. Inferential and
analogical reasoning touching reactions in consciousness
must be charily employed. When we are deahng with
fundamental evidence, it would be well to avoid analogical
reasoning altogether. Such at least has been my aim in
this book.

By introspection, we inferred the process of homo-
phony — the co-existence of the mnemic and the new
original excitations — from the mode of our recognition
and the perception of difference. In a reckless non-
scientific mood, we might say that a dog whipped once
recognises the whip, and manifests this by unmistakable
objectively-perceived reactions, thus clearly proving the
existence of homophony. This, however, would be alto-
gether wrong. The only thing that is quite certain is
the ecphory by a definite stimulus of specific engram-
complexes. The sight of the whip acts ecphorically on
a complex in which the engram of the sensation of pain
plays a prominent part. This is justly inferred from the
appearance of the corresponding reactions. It is true
that a mnemic process is involved, but it is by no means
certain that it is associated with the same feeling of
conscious recognition which we ourselves experience. It
would be necessary first to prove the identity of these
acts in consciousness, before one could infer the presence
of homophony in the lower creatures.

Rejecting this method of adducing evidence, let us
examine the objectively-demonstrable reactions for those
criteria by which may be proved the existence of homo-
phony, the unison of mnemic and new original excitations.

Using objective methods, homophony is more easily
proved in those cases where the mnemic excitation and
the new original excitation are not perfectly superposed.
For then reactions arise whigh can only be interpreted

156 THE MNEME

as reactions against the incongruity of the mnemic and
the original excitation. When playing with a dog which
takes great delight in recovering objects— fox terriers
are best for the purpose — we vigorously throw small
stones not easily recognisable when in rapid flight, the
dog, with muscles taut and head uplifted, intently watches
each movement of our arm and hand. The flinging
movement executed and the stone started on its flight,
the animal, quickly turning round, rushes off in the direc-
tion of the stone. We repeat the action several times.
We then make the same movement, but without flinging
the stone. At first the dog reacts exactly as before.
In the absence of the stone to seize and bring back, the
dog redoubles its attention. But after being deceived
several times, the animal focuses its attention more
accurately, with the result that the structural detail of
the original complex becomes thereby completed. The
reaction of turning round and rushing off in the direction
of the throw now takes place only when the dog sees
the stone actually flung ; that is, only with the perfect
congruity of the homophony of the mnemic and the
new original excitations. At the incongruity of move-
ment without the stone-throw, the animal reacts either
by an attitude of readiness to go, or, in its excitement,
with a false short start which immediately gives way to
the previous intense alertness. The different behaviour
of the animal can be rightly regarded as reactions which
vary according to the congruity or incongruity of the
mnemic and the new original excitations.

Another illustration touching the reactions which follow
the congruity or incongruity of homophony may be
taken from the realm of sport. Some animals, chiefly
among the higher mammals, may be allured by certain
sounds and tone sequences. In the absence of a special
gift for the imitation of animal voices, some sportsmen
make use of instruments which mimic the rut cry of the
stag or the sex call of the doe, in order to entice the
animals within the range of their guns. It has been noted
that, ceteris paribus, the reactions of the creatures vary

MNEMIC EXCITATION AND HOMOPHONY 157

according to the greater or less perfection of the mimicry
of the natural calls. Where the resemblance is not
perfect, the game, contrary to our expectations, is neither
indifferent nor unduly frightened. Rather does it take
an interest in the sounds, reacting to them similarly ai
to the genuine calls, thus proving that the mimicry has
acted ecphorically on certain engrams, although only to
the extent of similar and not identical reaction. By
not venturing too near the spot from which the sounds
come, the game shows that it has noticed a difference.
Of course, it is assumed that the game does not otherwise
scent the hunter. If, however, the call instrument is
used by an expert, and the sounds produced are so like
the natural cries that there is little perceptible difference,
the behaviour of the game changes. It then reacts to
the mimicry much in the same way as to the natural
sounds, the more readily the less often it has heard the
genuine calls ; for the younger it is, the less precise are
the engrams of that kind it possesses. But an old and
experienced roebuck, with all its pasha desires in full
vigour, reacts even in case of the most perfect mimicry
to the shght incongruity which still exists at the homo-
phony of the mnemic and the original excitations. It
is timid in its approach, it lurks among the trees, it is
strangely alert, and, as one on the stand can easily observe,
its behaviour is quite different from that which follows
when the genuine caUs of the doe reach its ear.

Numerous similar cases of manifestly varied reactions
on the imperfect homophony of mnemic and renewed
original excitations might readily be given. But perhaps
I may be allowed to conclude with an example where
the mnemic excitation involves the operation of inherited
engrams, and not, as in the previously cited cases, of
engrams in the main individually acquired. It is well
known that incubated birds without any experience of
the nest are able, when the time of mating arrives, to
build a nest and to finish it off almost as perfectly as
those members of their species who have already passed
through several periods of mating.

158 THE MNEME

This reactive manifestation of a succession of mnemic
excitations, for so we must regard it, is associated only
with the mating period, is absent in castrated animals,
and terminates with the completion of the nest. It
might be thought that the termination depended on the
passing of the internal energetic condition into another
phase, or on the exhaustion of the disposition through
the natural subsidence of the excitations, a condition
which is manifested in the cessation of the reactions.
Neither of these ideas meets the case. For we have but
to remove the completed nest, to elicit at once the same
succession of nest-building reactions. And this may be
done three or more times. On the other hand, by placing
an artificial nest at the disposal of the birds before they
begin building, we effectually block the series of reactions.
But the nest must be of a certain form, size, and con-
sistency in order to exercise any such influence on the
phaseogeneously ecphorised reactions of nest-building. If
the form differs greatly from that which is hereditarily
peculiar to the respective species of birds, or if the nest
is appreciably larger or smaller, or if the material of
which it is made is too hard or not sufficiently dry, the
bird, after a minute examination of the structure, will
either discard it altogether and proceed to the building
of a nest, or turning to the thing provided will entirely
reconstruct it, removing the unsuitable parts, and replacing
the missing ones. And all this is done by creatures that
have never seen a nest of any kind and possess no indi-
vidual experience of the rearing of young. Bees, before
they have had experience of the natural comb, manifest
a like behaviour towards half-finished artificial combs
which the bee-keeper may place at their disposal. For
example, they correct the deviations of the artificial comb
from the strictly perpendicular.

In all these occurrences, the normal course of the
reactions is modified by the original stimulus-complex —
here the provision of the artificial nest or comb — in a
manner which stands in a definite relationship to the
difference between this original stimulus-complex and

MNEMIC EXCITATION AND HOMOPHONY 159

tjie final effect of the mnemic excitation, that is, the
making of a specific kind of comb or nest. Or, as we
may phrase it, as long as the original stimulus-complex
of the artificial nest or comb shows appreciable incon-
gruities with the nest normally produced by mnemic
reactions, the organism manifests reactions which tend
to obliterate incongruity. When congruity has been
attained, the specific reactions cease until some inter-
ference again demands their exercise.

The objection may be raised that here is no case of
congruity or incongruity of homophony, because the
mnemic excitation evoked is undoubtedly an unconscious
one ; that is, no clear image of the final product of, say,
nest-building actions could possibly appear to birds
mating and hatching for the first time. It may be ad-
mitted that most probably no such clear image arises
in consciousness. But for the sure solution of such
problems of consciousness, we need criteria which in the
nature of things are unavailable. Such problems can
only rightly be discussed where the application of the
introspective method is possible. But it is valid to meet
the objection stated by proving that in ourselves the
presence and the evidence of the homophony are inde-
pendent of the existence of conscious sensations.

In intense mental occupation we are apparently in-
sensible to other impressions. The playing of hackneyed
pieces on the piano or violin in the next room does not
disturb us. We continue brooding over our problems
without deliberately listening to the sounds. But let the
player make a mistake or elaborate the theme or so render
it as to conflict with our mnemic knowledge of it, and at
once the incongruity is followed by the reaction of a
start or a smile or a frown, of which for a time we may
be quite unconscious. Prolonged duration of the incon-
gruity, however, invariably leads to conscious reaction.

Cases of homophony, in which neither the mnemic
nor the Original excitations rise into consciousness, occur
continually in daily life. We discover a new walk, and
on our second or third experience of it we are generally

i6o THE MNEME

conscious of the homophonies involved. But on further
acquaintance with the walk, we are no longer conscious
of these homophonies, as when we walk along deeply
wrapped in thought or engaged in absorbing conversation.
The existence of the homophonies can in the latter cases
be demonstrated only by reactions other than conscious
ones. This evidence is easily derived from the fact that
automatic, unconscious walking is possible only when
mnemic and original excitations actually coincide. The
ability to reach the destination without the co-operation
of consciousness is affected by the appearance of incon-
gruities, as, for example, the blocking of the path or the
disturbance of the roadway. Thus, in the life of every
human organism, unconscious mnemic homophony plays
at least as important a part as conscious homophony,
and the question whether or not an homophony becomes
manifest in the consciousness of an organism is of rela-
tively minor importance.

We conclude that the presence and action of homo-
phony become manifest to us by special reactions. In
experimenting with ourselves in the case of mnemic and
original excitations which rise into consciousness, the
homophony manifests itself by the sensory reactions of
recognition and of perception of difference. Where the
excitations do not rise into consciousness, the homophony
is known in an indirect way by the appearance or absence
of objectively perceptible reactions. The same criterion
applies to the demonstration of every homophony in
organisms other than one's self. For in these, homophony
can only be inferred from the appearance of objectively-
perceptible reactions, whose characteristics lie in modi-
fications dependent on the congruity or incongruity of
the original state of excitement with a state of excite-
ment that had previously been experienced, either by
the same organism or by its ancestors. For the ecphory
of this mnemic state of excitement similar conditions
must again prevail. With the simultaneous presence of
the corresponding original excitation, the mnemic state
of excitement can very readily be inferred by the external

MNEMIC EXCITATION AND HOMOPHONY i6i

observer from the reactions arising from a possible in-
congruity. But the most convincing confirmation of
the' presence and working of homophony is furnished by
those reactions through whose working the incongruity
or discord is resolved.

Sensory reactions can only be perceived by introspec-
tion. In regard to the quality of reactions other than

' sensory, we have so far given examples from those re-
actions resulting from the contraction of muscles. It is
clear that it is only by the reaction that we are able to
recognise excitations original, or mnemic. But the same
excitation may manifest itself by very different reactions.
Take, for example, protoplasmic movements, muscular
contractions, metabolism, and the phenomena of growth ;
the first two are motor reactions, the third are metaboUc
reactions, and the last are plastic reactions. Now, there
is no reason whatever why we should consider excitations
manifesting themselves mainly or exclusively by motor
reactions as a class different and separate from excita-
tions apparent by plastic or metabolic reactions. We may
take the following example : — Heliotropism and heUo-
taxis in plants and animals are phenomena traceable to
the stimulus of light on the organic substance. They
can be classified under heads which apply to all organ-
isms. But the reactions by which these stimulations
and excitations — say, in plants — are manifested may be
either motor, as in amoeboid and ciliary movements, or
may depend on osmotic processes, such as changes of
turgescence, or finally, may be plastic reactions, as in
curvatures of growth. No plant-physiologist studpng
the effect of light on the organic substance of the plant
will regard excitations, which manifest themselves by
osmotic reactions, as fundamentally different from those
which manifest themselves by motor or plastic reactions.
What we allow to the original excitations may equally
well be claimed for the mnemic excitations. In the
previous chapters we acted on this conviction, when, in
describing the plastic reactions of the fall and the fresh

sprouting of leaves, and the osmotic reactions of the

II

i62 THE MNEME

so-called sleeping movernents of plants, we regarded them
with equally good reason as manifestations of mnemic
excitations.

But the moment we recognise in any specific case the
presence of excitations, the presence of potential homo-
phony may also be inferred. To educe the evidence for
homophony, we have to show, first, that under certain
circumstances the conditions exist for the simultaneous
rise of mnemic and corresponding original excitations,
and secondly, that given these conditions, reactions
regularly follow which modify themselves according to
the congruity or incongruity of this potential homophony.
Among these reactions those which tend to regulate the
incongruity are the most striking. There is a large group
of plastic reactions which tend to remove any specific
incongruity between the normal developing or developed
stage and an actual plastic state already existing. These
reactions may be described as regulations and as re-
generations in the widest sense, that is, inclusive of the
conceptions of post-generation, reparation, etc. To justify
the inference of homophony from these reactions, first
the evidence would have to be adduced that in the given
cases the conditions exist for the simultaneous presence
of certain mnemic excitations with the new corresponding
original excitations, and that at the homophony of these
two excitations the reaction under consideration tends
to remove any incongruity that may exist.

The evidence required can easily be furnished, but to
avoid repetition, I reserve it for the following chapters,
which are devoted to the analysis of the mnemic factor in
ontogenesis, in regeneration, and in processes of regulation.

So far, in considering homophony, we have regarded
the mnemic state of excitement as something uniform
in spite of its complexity. The idea equally applies
when the mnemic state corresponds to the reproduction
of a single preceding excitation. But how if the mnemic
state reproduces a frequently repeated excitation ?

There are two methods for the solution of this question.
First, by the synthetic method we can note how the

MNEMIC EXCITATION AND HOMOPHONY 163

mnemic total increases at each repetition. Secondly, by
the analytic method we can set out the constituents of
a mnemic excitation originated by frequent repetition.

Let us begin with the second method, and endeavour
to analyse a mnemic excitation based on repeated en-
graphic action. We select a case of mnemic excitation
introspectively perceivable. Let us try, therefore, to
ecphorise the image of the bodily presence of a near
but absent relative. ,We are dealing with a purely
mnemic process. At first, the image recalled seems clear
and definite, but if it is that of a person with whom we
are in constant intercourse, we shall find on closer scrutiny
that the ecphorised image is, so to speak, generalised,
resembling somewhat those photographs which aim to
furnish the general character of a type by the super-
position of the pictures of different heads on one and the
same plate.

In the case under consideration, the generalisation
occurs by the homophonous action of different images
of a face which we have seen in many varied states and
situations, at one time pale, at another time flushed,
now serene, again solemn, once in this light and another
time in that. As soon as we inhibit the simultaneous
presentation of the multitude of images, and, on the
ecphory of the engram, focus only one definite engraphic
experience, the cognate mnemic excitation at once pre-
dominates over its congeners which faintly sound with it
in unison, and we straightway recognise the clear, sharp
outline of the face in some specific situation.

So, in the case of persons with whom we are in con-
tinual association, it is just this abundance of mnemic
excitations usually sounding in unison with each other
which, when their features are mnemically produced in
us, accounts for the vague outUne and generality of
aspect. But in the case of people we meet more rarely,
our attention during the mnemic reproduction of their
faces is usually focused on a particular experience when
the face made a special impression on us, and the engraphic
results were consequently well marked. By reason of

l64 THE MNEME

this accentuation, the features sometimes appear to us
more strongly deUneated than even those of our nearest
relatives, seen much more frequently and in the most
varying situations.

Where on the ecphory of a frequently recurring engram
no predominance of a single component results, that is,
of any one of the mnemic excitations sounding in unison
with each other, we may observe, so to speak, a growing
abstraction of the memory-image, similar, as we have
suggested, to the increasing vagueness of the contours
on the superposition of a number of exposures not exactly
corresponding to each other. The result is — at least
with man, probably also with higher animals — the genesis
of a kind of abstraction, which I call " Abstraction of
Homophony." Mnemic homophony can, without the aid
of any other mental process, furnish us with a kind of
abstract image of our friend X ; that is, an image robbed
of particularity of aspect and situation giving us X, dis-
sociated as it were from a definite moment in time. If
the sphere of the ecphorised engrams be further enlarged,
abstract images of a higher order arise. We can, for
example, summon up pictures of a white man or a negro.
I maintain that the primary formation of abstract con-
ceptions is based on such abstract images, and that this
abstraction, created in the above-mentioned way by
homophony only, is the precursor of the purely logical
process. It is no monopoly of the human species, for
we find it manifesting itself in different ways among all
the higher organised animals.

The fact that, by focusing attention on single homo-
phonic components, we can dissociate almost any one
of the. number from the rest proves that on the occasion
of the homophony there is no perfect blending of the
mnemic excitations.

The accuracy of our findings by the analytical method
may be tested by the synthetic construction of the results
of the frequent repetition of an engraphic action on the
basis of the general laws derived from our previous
investigations.

MNEMIC EXCITATION AND HOMOPHONY 165

In the case of the first ecphory of an engram by the
recurrence of the original stimulus, a solution is afforded
in the work already done. Let us describe the mnemic
excitation at its first ecphory as p^ (mn), and the original
excitation generated by the first repetition of the original
stimulus as p^ (or). Unisonant working follows, or, as
we say, homophony takes place ; but there is no blending
of these two excitations into one. The result may be
expressed as />, {raa.)-\-p, (or). Now, we have already
recognised the general law that, where two co-ordinated
excitations affect an organism, they are received and
fixed in co-ordinate engraphy. It is clear that there
can be no exception in the case of the excitations
pj^ (mn) and p^ (or), when respectively they lapse into a
state of latency. For if by the recurrence of the original
stimulus they are again ecphorised, they must manifest
themselves in co-ordination as an homophonous mnemic
excitation px (mn)-|-^g (mn), and in this form enter into
relation with the newly appearing original excitation
p^ (or). Thus there follows the threefold homophony
of pj^ (mn)-H^a (mn)-}-/'3 (or). A similar process occurs at
succeeding repetitions up to the Mth power. At the
(n+i) recurrence we have a mnemic homophony of the
mnemic excitations^, (mn)-l-^j (mn)-|-^j (mn) . . .p, (mn),
with the original excitation />„+, (or). The process
may thus be phrased : — At the ecphory of a combination
of engrams, owing its origin to frequently repeated en-
graphic action, what is given is not a single indissoluble
blend of the mnemic excitations — " coalescence," some
physiologists call it — but a unisonant chorus in which
the single components of an apparently uniform com-
bination of engrams, distinct indeed from each other as
to their time of origin, may be individually discerned.
Further, we have to remember that in most cases the
single components differ considerably from each other ;
for it will only rarely happen that an original stimulus
at its repetition resembles its predecessor in all par-
ticulars, and further, the energetic condition of the
organism itself suffers continual change. The original

i66

THE MNEME

stimulation, therefore, enters into association with various
complexes of engrams, and is, therefore, simultaneously
and successively differentiated from its predecessor.

Both analytic and synthetic investigations lead us to
the same result, that at each ecphory of a combination
of engrams which has been created by repeated stimu-
lation, there emerges a unisonant chorus of individual
componeiits, each of which is the resultant of a separate
stimulation.

With this understanding, we are better able to discern
the nature of those bifurcated successions of engrams
which can only be ecphorised alternately.

Let us take again the case of the two versions of the
line in the Ruhdiydt of Omar Khaj^Am, and assume
th.at we have heard both versions recited three times.
If we indicate by letters the single engrams generated
by the spoken words, and add to each letter-sign the
figure index corresponding to the number of its repetition,
we obtain by considering the succession of the first nine
engrams the following diagram: —

19 8 4 6

8

7 8 S

/

A' «»

/ /

fll _ fcl _ c' — d» — «'

_-yx

— g"- h' t's

a' — h' — d> — do — e'

-f

— if hi
/

aS — b3 — c3 — d3 — es

-P

-g^

a* — b* — c* — d* — e*

-f*

-g*
\

a5 — bi — cs — di — «s

-P

\ \

a* — 6* — c6 — d* — «6

-f

— ^ IJ' &*
>

MNEMIC EXCITATION AND HOMOPHONY 167

On ecphory, the mnemic excitation in phase i consists
of the homophony of the excitations a'-", or in phase 7
of the homophony of the excitations g'"*. The alter-
native arises in phase 8, when either the engrams A'-»
or the engrams ij*"" may be ecphorised. But why this
alternative ? Because, as has been shown at length in
the Mnemic Sensations (p. 356), a simultaneous mani-
festation of the branches of a dichotomy which consists
of word-engrams is altogether impossible.

Very different results arise where the possibihty exists
of a simultaneous manifestation of the two branches of
the dichotomy, as in the case of a piece of music which
begins as a simple melody, but develops into a duet.

Phaw 1

8

S

4

6

c

d —

'<

g —
1

d

C

1

a —

But if the simultaneous manifestation of the two lines
of reactions is impossible, the resulting alternative implies
either the total suppression of the ecphory of one branch,
or a jumping of the ecphory from one branch to the other.
In the latter case, where the reactions seem to move in
reciprocal succession, what may be called " mixed re-
actions " result. We are dealing with such a " mixed
reaction " when we combine the two alternatives, " in
a noose " and " with a shaft " as " The Sultan's Turret
in a shaft of light."

Such mixed reactions are not rare in the manifestation
of the individually acquired Mneme, nor are they alto-
gether absent from the manifestation of the dichotomic
successions of inherited engrams. We shall deal^with the
matter at greater length in the chapter on the import-
ance of alternative dichotomy in ontogeny. At the
same time it should be noted that in most cases of in-
hefrited, as well as of individually acquired, alternative

i68 THE MNEME

dichotomies, the ecphory takes place along either one
or the other path. In the case given on page i66,
either the homophonous engrams A'"^ or ij*"* become
ecphorised. The emergence of the mnemic alternative
depends on the ecphoric predominance of either h^~^
or Tj*"*, and the most diverse factors can turn the scale
to this or to that side. In the majority of cases, the
predominance of one of the two branches is assured by
our experience of the branches being quantitatively
unequal. If, for example, we have heard the second
version of the Omar KhayyAm line more frequently
than the first, then, ceteris paribus, the ecphory along
the path of the second version will predominate. Pre-
dominance is also affected by the recentness of the ex-
perience, the advantage lying with that path along which
the fresher experience has moved. It is for this reason,
as we shall learn later on, that reversion to atavistic
paths is usually avoided in morphogenetic cases. Further,
newly occurring stimuli of various kinds may lend pre-
dominance to this or that ecphory, and thereby rob the
other side of its apparently established predominance.
A reciter may know both versions of the Omar Khayyam
poem, but is accustomed to recite the second version.
If at the point of bifurcation we prompt him with the
word belonging to the first version, we may succeed
sometimes, by nieans of this original excitation, in leading
him into the unaccustomed path. Under certain circum-
stances, deterrent influences affect the ordinary process
and divert the course of the ecphory from the more usual
channel. The chapter on morphogenetic dichotomies deals
with this side of our subject.

It is evidently impossible to elaborate a general formula
by whose application we could ascertain beforehand the
path by which the mnemic alternative will progress in
every individual case. But in many cases, after the event
we can discern those influences which at the alternative
gave the preponderance to this or to that ecphory.

In conclusion, it may be pointed out that our examina-
tion of homophony has given us an insight into the

MNEMIC EXCITATION AND HOMOPHONY 169

peculiar relation of the repetition of a stimulus to its
engraphic action. It might be imagined that by one
doubly powerful stimulation we should obtain an engram
of effect equal to two ordinary stimulations. But most
of us know that in general this is not the case. In learn-
ing by heart, for instance, we get on more quickly by
superficial scanning oft repeated than by a single intense
and concentrated effort. The witchery of repetition in the
mnemic domain might be demonstrated by many other
examples. It becomes intelligible to us by the concep-
tion that the engraphic result of repeated stimulation
differs in principle from the result induced by a single,
but correspondingly stronger, stimulation.
The fundamental difference lies in the fact

THAT repetition OF A STIMULUS DOES NOT STRENGTHEN
AN ALREADY EXISTING ENGRAM, BUT GENERATES A NEW
ENGRAM, AND THE MNEMIC EXCITATIONS RESULTING FROM
ANY SUBSEQUENT ECPHORY OF THESE ENGRAMS ARE IN
HOMOPHONY.

PART III

CHAPTER VIII

EVIDENCE OF THE MNEMIC FACTOR IN ONTO-
GENETIC REPRODUCTION

Taking any organism whatever and surve3dng its onto-
genesis, no special reason seems to exist for regarding
the ontogenetic development as due to anything but
to the action of original stimuh, producing original ex-
citations. Fertilisation, we might say, acts as an original
stimulus determining the reaction of the first nucleus
and cell-division. The stimulus of position, resulting
from the juxtaposition of the first two cleavage-cells,
liberates, as an original stimulus, the next segmentation,
and so on. At a certain stage of development in Verte-
brata, for example, the free end of the optic vesicle
touches the ectoderm. The contact acts as an original
stimulus, it may be through a specific thigmomorphosis,
and this determines a plastic reaction, namely, the lens-
invagination of the ectoderm.

All the processes involved can be considered as effects
of pure original excitations. Although we are still far
from being able to analyse the series of internal changes
from the first alteration caused by the stimulus until
the final reaction, nothing here seems to point to the
action or co-operation of mnemic excitations. For the
repetition as such, characteristic of ontogenetic develop-
ment throughout the generations, does not require the
aid of mnemic principles for its elucidation. The validity
of our experience is largely dependent on the axiom that
hke causes produce like effects ; or, as we may say,
given certain premises, certain conclusions follow. On

173

174 THE MNEME

this principle, the repetition of ontogenetic facts can,
without any aid from mnemic processes, be explained
as we explain recurring natural phenomena, such as the
procession of the seasons, the alternations of the tides,
and the intermittent eruptions of geysers.

One feature, however, is characteristic of all these
repetitions. Much the same conditions must rule, in
order to secure the recurrence of the phenomena. In
the mechanical world, cases may be devised where the
one effect follows either from a sum-total of causes, or
from a fraction of this totality. It amounts to the same,
whether the opposing forces playing upon a wagon are
in the relation of five and four horse-power, or four and
three horse-power. The resultant is the same, a pull of
one horse-power in one direction. In this instance two
factors which already neutralised each other may be
eliminated. The resultant of a fraction of causes is
equivalent to that of the sum-total only when the eUminated
factors cancel one another. But where this is not the
case, the effect of a totality of causes is nearly always
different from that of a fraction, especially if the latter
be selefcted at random.

There is but one group of phenomena which forms
an exception to this rule, namely, mnemic excitations
with their concomitant reactions. Our previous investi-
gations have furnished evidence that one essential dis-
tinction between original and corresponding mnemic
excitations hes in the fact that the former are generated
by a definite original stimulus-complex, while the latter
may be roused by any one constituent of that complex.
To refer once again to our original example : Capri, the
barrel-organ, and the oil smell had to be present as original
visual, auditory, and olfactory stimuU and to act simul-
taneously, in order to generate the corresponding original
excitation-complex. The ecphory of the corresponding
mnemic complex, however, could be induced at any time
by the recurrence of any one of these stimuli. Stimuli
of warmth, moist air and presence of water induce large
broods and premature parturition in Sulamandra atra,

EVIDENCE OF MNEMIC FACTOR 173

an animal which under normal conditions produces two
metamorphosed young ones. With each repetition, the
effective strength of the stimuU on the individual increases.
In the end, the stimuh may be ehminated entirely, and
yet, at gravidation, premature parturition of large numbers
will take place. The induced change is manifested also
in the next generation. Mutatis mutandis, the habitual
and inherited parturitions of Salamandra maculosa are
equally affected, but in an opposite way.

In this last example we have an embryological pheno-
menon, which the most rehable criterion for these pro-
cesses, namely, the experimental examination of engraphic
action, proves to be undeniably mnemic. The same can
be said of the ontogenetic phenomena of the colouring
of butterflies described by Standfuss, Fischer, and their
adherents ; and also of the rich accumulation of results
in experimental breeding, which we owe to Blaringhem,
Bordage, Chauvin, Kammerer, Klebs, Pictet, Przibram,
Schroder, Sumner, Tower, and many others.

For the great majority of ontogenetic phenomena, an
experimental test of all potential engraphic stimulations
is impossible. We can, however, by experiment adduce
evidence that compels us to regard these phenomena as
mnemic. We may eliminate in a fairly arbitrary manner
one or other part of the conditions, and yet the course
of events will, at first, be altered only in so far as the
interference renders impossible certain reactions. A cell
that has been removed by the surgical knife cannot of
course respond by reaction, but as long as the reacting
organs are still there, the course runs on in a comparatively
undisturbed manner, in spite of the restriction of certain
conditions and the absence of some contributing elements.

Experimental evidence of this kind has already been
furnished for the ontogeny of the Metazoan groups. It
is obvious that the conditions of a course are greatly
altered by the removal of, say, one-half, three-quarters,
or seven-eighths of the system in which the course generally
completes itself. The alteration of conditions is particu-
larly effective when the course is determined, not by

176 THE MNEM'E

the influence of causes acting externally, but by changes
taking place within the system itself. An example of
this, latter condition is given when by experimental
interference we alter the conditions of an embryo during
its ontogenetic development. We may take a Ctenophore
when in segmentation it has reached the stage of eight
cells and divide it into two, four, or eight parts. No
regeneration takes place, but the two, four, or eight
sections continue their course of development almost
as if no such extraordinary change of conditions had
taken place. The Ctenophores being pelagic animals,
the larvae, perfect or imperfect, under the unnatural
conditions of the aquarium, perish immediately after
passing through their metamorphosis. We are, therefore,
unable to tell how much longer under more favourable
conditions the development of these mutilated part-
systems might have continued.

In much the same way, by mutilating the developing
eggs of Echinoderms, Annelids, Ascidians, Molluscs, etc.,
serious changes of the conditions may be effected without
any immediate essential alteration of the course of develop-
ment within the remaining parts of the system. It is
true that striking aberrations from the normal course set
in later ; but the point need not engage our attention
at the moment, as, in view of its great importance, we
shall presently consider it more minutely.

It has been stated that experimental mutilation can
be effected in a fairly arbitrary manner without disturb-
ing the rest of the system. But in this respect the range
of our freedom has its limits, which vary according "to
the species to which the organism belongs, and the stage
of development it has reached. A fuller discussion of
the matter will be found in Chapter XL

We have already found it characteristic of mnemic
phenomena — that is, the manifestations of mnemic ex-
citations by specific reactions — that they require for their
ecphory but a fraction, arbitrarily chosen it may be, of
the conditions which were required to generate the corre-
sponding original excitations. Jt is clear that ontogenetic

EVIDENCE OF MNEMIC FACTOR 177

phenomena resemble the nmemic phenomena in so far
as in their production fairly large and arbitrary subtrac-
tions may also be made from the conditions which normally
obtain. It is useful to point out that this is a general
characteristic of ontogenetic processes in all kinds of
organisms, and at all stages of development. But the
kind of alteration we may effect in the conditions, with-
out annihilating or arresting the course of development,
varies according to species and stage of development.

In the animal kingdom, the number of classes in which,
after experimental interference, the course of develop-
ment continues for a time as if no such change had been
effected, is limited. But with the Hydromedusse, Amphi-
oxus, Teleostei, and Amphibia, the course is modified
almost immediately after the change has set in. The
nature of this modification is, however, of such a kind as
to furnish a strong argument for the mnemic character
of the excitations which manifest themselves in the plastic
reactions of ontogenesis. Modification of the course often
follows even in those cases where, for a time, the
development continues as if no change of condition had
set in. This, for example, holds good in regard to the
Echinodermata.

In what does this modification of the course consist ?
In reference to the elimination by experimental inter-
ference of a part of the physical conditions, we used the
phrase " undisturbed course." The word " undisturbed "
relates only to the reactions of the remaining portion of
the system. It cannot, of course, refer to the reactions of
the eliminated parts, for it is obvious that, with the ex-
cision of parts, the reactions appropriate to them fall
away. But as regards the parts that remain, we note
that, after some httle time, or even in rare cases jointly
with the cycle of the usual reactions, new plastic reactions
set in, which, however diverse in manifestation, have
this in common, that they tend to effect a re-establish-
ment of the conditions disturbed by the interference.
Or, as we may phrase it, they finally establish a congruity
between the state — at the moment we refer specifically to

12

178 THE MNEME

the morphological state — of the remainder of the system
and that state which the whole system would have
reached if no interference whatever had taken place.

During the analysis of mnemic homophony, in the
preceding chapter, we noted certain reactions which
tend to abolish a specific incongruity between two states
— the original state of excitement, and the corresponding
mnemic state. At the moment, however, we are deaUng
with morphological states. If, therefore, we wish to
establish a direct relation between the above-mentioned
ontogenetic observations and the results obtained during
the analysis of the mnemic homophony, we shall have
to prove — ^first, that with the occurrence of ontogenetic
phenomena, it is valid to infer states of excitement from
morphological states ; secondly, that conditions exist
for the presence of an original state of excitement as
well as of a corresponding mnemic state of excitement ;
and thirdly, that new reactions either partially or entirely
effect the removal of incongruities in the homophony of
those two states of excitement.

The question, whether states of excitement correspond
to the morphological states clearly observable at onto-
genesis, is not difficult to meet. But in answering it,
we shall take into account the whole body of morpho-
logical states, including those of the developed organism
on the completion of its ontogenesis. It is evident that
the energetic condition of a system, that is, its internal
energetic condition, is partly determined by the morpho-
logical state. Only partly determined, for besides the
morphological state, chemical, thermic, electrical, and
other states play, of course, an important part. The
various elements of the energetic condition, together
with the morphological state, determine the respective
organic state of excitement ; and the original excitations
so generated are joined by mnemic excitations, which,
at that moment, are ecphorised in the organism; The
morphological state of an organism, therefore, is only
in part responsible for the extremely complex state of
excitement, which, at a given moment, may develop

EVIDENCE OF MNEMIC FACTOR 179

in the organism. But although the morphological state
is only one of several factors, its importance cannot be
overrated, for never for a moment does it cease to act.
With its changes the state of excitement must also change.
A part of the complex state of excitement, therefore,
definitely depends on the morphological state of the
organism. This we shall define as the morphogeneous
part of the excitation-complex.

This distinction of a morphogeneous part from the
whole of a simultaneous excitation-complex could hardly
be supported by closer analysis, but it facihtates our
understanding of the subject, and may, therefore, be
allowed to stand as a temporary expedient in the present
phase of our investigation. If we consider the morpho-
geneous part of a simultaneous excitation-complex as
the sum-total of the excitations set up by " stimuli of
position," we must not forget that the reference to its
contents is but summary. On fmther analysis we shall
find that these stimuli of position are resolvable into
various classes. For our present purposes, however, the
collective definition given will suffice.

To the first of the above three questions, then, we
make answer that to the morphological state of a developing
or fully developed organism there corresponds a definite
part of its state of excitement. This we have summarily
described as the morphogeneous part of this state.

The second question asks whether the conditions exist
for the presence not only of an original but also of a
corresponding mnemic excitation in those cases where
the conditions have been changed by experimental inter-
ference, and where a modified coinse of development
has followed.

The conditions certainly exist. Two things are required
for the rise of a mnemic excitation — an engram and its
ecphory. As the engram in this case has to produce
an excitation corresponding to the morphogeneous area
of excitement, that can only be if the engram is the
result of the repeated action of a similar morphogeneous
excitation. It has been made clear that, in cases of

i8o THE MNEME

ontogenetic process, the stimuli of position produce, in
each successive developmental stage, excitations which for
the individual in question are a singular non-recurring
experience. But the same or very similar complexes of
stimuli have produced Uke excitations in the innumer-
able ancestors of this individual. Our contention is that
these influences have acted engraphically, and that the
resultant engrams have been transmitted to the offspring.
In the nature of things it is impossible to demonstrate
this by the direct experimental re-creation of these
engrams in the manner, say, of the experiments elabor-
ated by Kammerer, Chauvin, Standfuss, Fischer, Tower,
Bordage, and others. Our task is to prove the mnemic
nature of definite excitations on the occurrence of specific
states (see pp. 65, 66) . For the moment we shall regard
what has to be proved as proven, and assume that the
morphogeneous excitations have acted engraphically in
each generation, and that the engrams have been trans-
mitted to offspring. The question then would be whether
the conditions exist for the due ecphory in the offspring
of these morphogeneous engrams thus hereditarily trans-
mitted.

In each generation the morphological states whose
" energetic " action produces the morphogeneous engram-
complexes form a continuous succession. The rational
inference is that the morphogeneous engram-complexes
are successively associated, and that the ecphory of
the first in a sequence effects the ecphory of the related
series of engrams. Granted, therefore, at the beginning
of ontogenesis the ecphory of the first engram — a fact
which we shall prove later — the condition is thereby
given for the due ecphory of the successive engrams.
The mode of this we shall investigate presently.

There is a second operative condition which depends
on the ecphory by original stimuli of individual engrams
in the engram-succession. This second condition also
we shall examine later (p. 183). At the moment let us
concern ourselves with tho^e concrete cases, from which we
Started, of ontogenetic development altered by experiment.

EVIDENCE OF MNEMIC FACTOR i8i

We noted that the removal of parts of the developing
organism was followed, ' generally after some time, but
in rare cases immediately, by new plastic reactions in
the remainder of the organism. These new reactions,
different from those usually given by the corresponding
parts of the organism, varied greatly according to the
specific case. But they all had this in common, that
they ultimately effected a re-estabhshment of the con-
ditions thus disturbed ; in other words, they estabUshed
a congruity between the morphological state of the re-
mainder of the organism whose development still continued
in spite of the interference, and the morphological state
of that stage which the organism itself would have reached
if no interference had taken place, a stage through which,
of course, its ancestors, unaffected by experiment, naturally
passed. This latter state we know from the study of the
normal ontogenesis.

Now, we recognised that the morphological state of
the organism after operation is a factor determining
a corresponding origineil excitation in the affected
organism. But to the morphological state which the
organism would have reached in the absence of experi-
mental interference corresponds the mnemic morpho-
geneous state of excitement whose presence so far we
have assumed, but for whose release the necessary
conditions have now been shown to exist. It is clear
that this mnemic morphogeneous state of excitement
has not been altered by the experimental interference ;
it belongs to the accumulated stock of inherited
engrams, which, as we have suggested on page ii6,
and shall definitely prove in Chapter XI, is allocated
in equal measure to each mnemic protomer of an in-
dividual, and therefore cannot be destroyed by the
removal of morphological sections from the nexus of
the whole.

The expression — " the morphological state which the
organism would have reached if no • interference had
taken place " — acquires its true significance only when we
realize how dependent this state is on the corresponding

i8a THE MNEME

mnemic excitation actually present in the organism, after
as well as before any experimental operation.

Disregarding the two morphological states, one of
which cannot of course possess any reality in these cases,
and considering only the actually existing states of excite-
ment, we conclude that plastic, regulating reactions,
which at the disturbance of development manifest them-
selves apart from the usual reactions that continue the
development, fall under a category of reactions already
known to us. We showed in a preceding chapter (p. 155) ,
that by the objective method mnemic homophony can
be inferred only from the presence of reactions which
are objectively perceptible, and whose characteristic is
" that they modify themselves strictly according to the
congruity or the incongruity of an original state of excite-
ment with a state of excitement that has already existed
in the same organism or in its ancestors, and for the
ecphory of which in its aspect as a mnemic excitation
the necessary conditions again exist." The strongest
evidence for the presence and effectiveness of homo-
phony is furnished by those reactions which tend to
remove the incongruity (p. 160).

Our inference from the known facts can be succinctly
stated. On experimental or casual disturbance of the
ontogeny, reactions arise whose nature and scope are
determined by the incongruity existing between a morpho-
geneous original excitation and an ancestral morpho-
geneous excitation, that is, one that may be found in the
direct ancestral line of the organism. These modif3dng
reactions in the course of time remove the incongruity
relating to the excitations. From these reactions we
infer homophony, and conclude that simultaneously with
the morphogeneous original excitation the ancestral
morphogeneous excitation has reappeared as mnemic
excitation.

Before passing to the detailed consideration of ontogeny
as influenced by this homophony, we may be allowed to
revert to a point previously mentioned.

We saw that the morphogeneous engram-complexes are

EVIDENCE OP MNEMIO FAOTOR 183

successively associated, so that the ecphory of the first
engram effects in sequence the ecphory of the entire
engram-series. But we added at once that there existed
a second possibility, in that any one of the individual
members of the series might be ecphorised by original
stimuli. Let us consider this latter possibility more
closely, and from the starting-point of an active mnemic
morphogeneous excitation-complex trace the related series
of excitations and morphological changes which come
into activity in any ontogenetic course. This excitation-
complex is manifested in a body of reactions, and the
energetic condition thereby created acts as an original
stimulus, generating an original excitation-complex which,
in the case of an undisturbed ontogenesis, corresponds
on the whole to the mnemic excitation-complex from
which we started. The process is illustrative of the
congruity of the Homophony. It is clear that in this
case each of the two homophonous excitation-complexes
acts ecphorically on the succeeding associated engram-
complex (see diagram I on page 184).

We get an idea of their joint action in the introspective
examination of an analogous case. When a well-known
tune is played to us on any instrument, mnemic and
original excitations together act ecphorically on successive
engram-complexes. If suddenly the plapng ceases, the
mnemic course for a time still runs on. When the
latter comes to a stop, a fresh impulse may be derived
from a few bars played anew. As regards the tempo
of the successions, either the original or mnemic excita-
tions may predominate. But here, too, a congruity of
the homophonies will in time become estabUshed. In a
conductor, for example, who allows his orchestra to carry
him along with an accelerated tempo, the original excita-
tions predominate over the ecphorised mnemic excitations.
Another conductor, however, whose mnemic excitations
may possess relatively greater strength, overcomes the
force of the original excitations, and by the restraining
influence of his baton subordinates these to the mnemic
forces. In each case we see the ecphoric action of both

i84

THE MNEME

H

* ft

o .3

ft

a

bn

8

Ei

1

1

0)

^

la

■| a'
i °
I"

o o ^-.
■■C-Eb^

U ft c4

•rt " ■'-'

d
S o

s

J) -^

<U u

o

I"

•A

>> iH e
ij tuorfa 53

.2, <a
A

d nl
o o ,-,

•■a -a -a.

o ft rt

la"
ft

(U O

a

s-a

g-a-s-

ft

a

a a^

o

11

g ri

-So"

I «

.4

EVIDENCE OF MNEMIO FACTOR 185

I

.2. -£.„

00.

SjB

<»

u

> lU ^

o ■&
g

a

'a-
o

•3

o

a
■&
•c

o

"60 •«

fl O B
u

1:1

O

!U

>v

o "d

« S "

P^ ^rt CO
ft g

(D 2

O

•a

o

I-
-t-> "

H « _S

a ^

.3 «

••QO.

■^

43

0)

i— <

cd

tin

1

u

fr

q

00.

X

!

u

•I

I

•d

60

as

S3 9

■as

-3 s

c4

a

•c

o

a ••
>

■if

.. tj <*

1-1 .a .« -M

O 60 O 2
Q .9

s';

■I

g

+>

'S

n

o

g"3

*^ 60
" 5

■■g-2

IS o

|.a tt

i86 THE MNEME

kinds of excitations. For their action the rule holds
good that where, in the homophony, congruity does not
exist, it will ultimately be estabhshed in one way or
another.

During ontogenesis, the tempo of the course of succes-
sion is predominantly governed by the tempo of the
original stimulus-complexes, which in their turn depend
on the tempo required by the plastic reactions. For
while the tempo of the mnemic courses may easily be
accelerated or retarded, that of the original excitations,
depending as it does on plastic reactions, is determined by
morphological processes which are influenced by various
external, mainly thermal, conditions.

During ontogenetic development the ecphory of an
engram results from the joint action of the antecedent
mnemic and of the homophonous original excitation-
complex. This statement is not greatly affected by the
fact that the original excitation-complex may be mutilated
in consequence of experimental interference. For it has
already been eni;nciated as a general mnemic law that a
partial recurrence of a stimulus-complex acts ecphorically
on an engram-complex.

We will refer later to specific cases of the ecphory of
ontogenetic engrains. At this point, however, it may
be well to insert a couple of diagrams, one (I) of which
represents the normal ontogenetic course, and the other
(II) the course when the ontogeny has-been disturbed
by experiment. In this latter diagram only the courses
of such excitations as manifest themselves in plastic
reactions are considered.

It is obvious that the complicated assemblage of natural
phenomena can never be adequately represented by a
mere diagram. The utmost we can expect is, by more
or less arbitrary distinctions, omissions, and summaries,
to simplify the main workings and to present valid
generaUzations. Our diagrams are the result of this
simpUfying process.

The division of ontogeny into phases, for example, is
a purely arbitrary proceeding, whether the duration of

EVIDENCE OF MNEMIC FACTOR 187

the phases is measured by sidereal time, or by a standard
derived from the organic course itself. If, in some definite
ontogenetic process, say that of a total symmetrical
egg-segmentation, we choose the period required for
the completion of a nucleus-division as the time-unit
for the phase-division, we cannot be sure of our conclu-
sions, since this period is by no means constant for the
organism, but varies according to external and internal
conditions. From nucleus to nucleus the duration varies
within fairly wide limits. For practical purposes, however,
it may be allowable to take the mean time between
the nucleus divisions, and apply it as a standard for
the phase-division of the segmentation of the organism,
so long as the arbitrariness of this break in continuity
is borne in mind. We need not give to the phase-division
a deeper significance than, let us say, to the division of
a melody into bars.

In a piece of polyphonic music a good many difiEerent
successions may within the bar move in very diverse and
distinct tempos, and the continuity of the single com-
ponent tones may be broken within the bars themselves
as well as between any two bars.

In our diagrams I and II the compUcated successions
range themselves in their totaUty with apparent exactitude
into divisional phases. This, however, must be regarded
as an arbitrary simpUfication. In concrete cases, such
an orderly and regular arrangement is practically im-
possible. That excitation-complex a acts ecphorically
on engram-complex b, is, of course, but a summary mode
of expression. With regard to the connection of the
engram-complexes with each other, the reader is referred
once more to the general exposition on page 95, and to
the diagram on page 97. What is said there regarding
the intimate and the distant connections of the single
components of the simultaneous and successive engram-
complexes is valid also for the connection between the
components of the ontogenetic morphogeneous engram-
complexes. Thus, in a simpUfied diagram after the
manner of the one on page 97, the composition of the

i88

THE MNEME

morphogeneous engram-complex during a symmetrical
egg-segmentation would appear as below.

In this diagram we see that the connection of the
engrams presents dichotomies subject to simultaneous

t^^adeiCl

ecphory, the successive connections of the engram-complexes
being in the case given more closely related than are the
simultaneous connections. This follows from the fact
that in ontogenesis the simultaneous complexes are less
coherent than those of a lineal succession.

EVIDENCE OF MNEMIG FACTOR 189

This peculiarity may be thus explained. The ecphory
of certain components of a simultaneous engram-complex
may be either accelerated or retarded by diverse occur-
rences, and consequently may take place in an earlier
or later phase than that of other components in, the
simultaneous association. This is impUed in the estab-
lished fact that the developmental maximum of the
organ ordinarily fluctuates in the single " stages " within
certain relatively narrow limits. The composition of
the simultaneous complexes at each recurrence in the
succeeding generations varies within a definite margin.
It follows, therefore, that this fact should be mnemic-
ally expressed in a blurring of the ecphoric strength of
the simultaneous associations. On the other hand, the
ecphoric force of the successive associations, by reason
of their practical identity each time they recur, increases
with every new recurrence, thereby far surpassing in
effectiveness the less coherent simultaneous associations.

We may compare the ontogenetic course in this respect
with the reproduction of a piece of music, where each
separate part is perfectly performed, but where the har-
monic working of the parts is subject to slight vacillations,
which are yet insufficient to affect the general success
of the performance. People of limited executive ability
do this sort of thing on the piano when the rhjrthm of
the bass varies from that of the treble, or when, in accom-
panying their own singing, voice and accompaniment
take relatively independent tempi.

When the deviations during ontogenetic development,
as also in musical reproduction, pass a certain point,
readjustments are effected by various reactions which
operate when the incongruity between the original and
the mnemic excitations exceeds a certain limit.

I shovdd hke to call attention to one more fact. It
is the attainment of a definite state and not the mere
number of developmental steps which acts on a certain
engram as a successive ecphory. As pointed out in
Chapter III (p. 75), the beginning of gastrulation in
the Echinoderms does not depend simply on the number

igo THE MNEME

of the preceding cell-divisions, but follows primarily on
this specific reduction of the size of the single cells. With
complete ova, the gastrulations begin after a relatively
great number of divisions ; but with ovum-fragments
or with isolated blastomeres, the necessary number of
preceding divisions is less, proportionate to the size of
the cells. If two developing germs are forced to unite,
as they were by Driesch, into a single new individual
of double size, then before gastrulation sets in, the number
of divisions is necessarily greater than in the case of
ordinary complete ova.

There is still one more point whose consideration may
obviate possible error. On page i88, we gave a dia-
gram representing the composition of the morphogeneous
engram- or excitation-complexes during symmetrical egg-
segmentation. If we now elaborated a second diagram,
giving not the connections of the engramsjar excitations,
but the successive connections of the cells, it might be
inferred from the concurrences of these diagrammatic
representations that, in each phase, one engram or exci-
tation-component corresponded to one definite cell in
such a way that we might think of the component as
localised in this cell. But the inappropriateness of this
notion has already been indicated in the chapter which
dealt with the radiation of the excitations over the
whole organism (p. 123). Further, the idea could only be
validly conceivable if each cell in regard to stimulus-
conduction were isolated from its neighbour, which is
certainly not the case. It is therefore impossible to
say that each morphogeneous excitation-complex runs
its course within the organism according to a strict mor-
phological localisation, and that one excitation-component
runs off in this cell, and another in that. So definite
an allocation of excitation-complexes to specific areas
cannot be accepted, for every mnemic protomer in the
organism is influenced by the morphogeneous excitation-
complex as the latter runs its course.

CHAPTER IX

THE INITIAL ONTOGENETIC ENGRAM AND THE
ONTOGENETIC COURSE OF DEVELOPMENT

The evolution of a series of organisms presents itself
as a continuous line which passes through phases of
time and space. To each phase in time an individual
corresponds. The individual may be regarded as a
phase in space. While the continuity in time is an
altogether uninterrupted one, the continuity in space
may be broken, since at the beginning of each phase of
individuaUty a, separation in space usually takes place.
This separation is the rule in sexual propagation, but it
may not occur in asexual propagation, or, if it does, only
comparatively late. In spite of this break in continuity,
the evolution proper represents, without exception, a
continuous hne, whose interruptions are of a secondary
nature ; that is, they occur at a point already passed
through by the leading developmental Hne.

This break in continuity, however, is of special importance
for our present consideration, because thereby a mnemic
separation between the parental and the filial organism is
effected. It is only on the completion of this separation
that the possibility arises of the filial organism acquiring
engrams in which the parental organism has no share. By
such a separation it acquires its own mnemic individuality.

But although the filial organism may detach itself from

the parental organism, it does not follow that in the majority

of cases the new ontogenetic cycle begins straight away.

Ignoring the male germs, which from other and external

causes may be imable to commence the ontogenetic cycle,

and considering only the more favourably disposed female

in

192 THE MNEME

germs, we find that these, after the maturation changes,
are in most cases unable to start the new cycle of develop-
ment without the aid of an external impulse. In the absence
of this impulse, the ovum ordinarily remains for a time in
a state of rest, and then gradually perishes.

This maturation stage, as is usual in descriptive and
experimental embryology, may be regarded as the starting-
point of the ontogenesis. We note that in most cases a
specific external stimulus is required to effect the ecphory
of what we may call the initial ontogenetic engram, and to
arouse the mnemic excitation a (see the diagram on page i88).

The stimulus that normally acts ecphorically on the initial
ontogenetic engram is connected with the process of fertilisa-
tion. Interesting and important as the matter is in itself, we
will not now examine to which of the numerous energetic
influences generated by this process this ecphory is due.

But it is particularly instructive and strongly indicative
of the mnemic nature of the excitation initiating the onto-
genesis, that the liberating stimulus may be one of many
different classes of stimuli, some of which are able to induce
a parthenogenetic development, that is, without fertiliza-
tion by the male element. The reader is referred to what
- was said on pages 45, 70, on vicarious ecphory. Loeb, who
dealt with this problem in a series of excellent experiments,
sees the first decisive step in ontogenetic development in
the formation of the membrane of the egg.

He effected membrane-formation by the specific action
of formic, acetic, propionic, butyric, and other fatty acids.
Mineral acids, such as hydrochloric acid, were also effectual.
He also found that saponin, solanine, digitalin, the bile
salts, the specific fat-dissolving carbohydrates such as
amylene, benzol, toluol, and chloroform, ether and the
alcohols, the presence of free oxygen, and, under certain
circumstances, a mere raising of temperature, would cause
the formation of the membrane. The same end was attained
by the injection either of blood from animals belonging to
a different family, or of extracts from organs belonging to
animals of different species. In a great number of these
cases it happened, however, that the development thus

THE INITIAl ONTOGENETIC ENGRAM J93

started ran off abnormally and came prematurely to a
standstill. This was most probably caused by an injurious
substance produced during the course of the abnormal
development. Loeb discovered two ways of counteracting
the effect of this poisonous substance. He treated the egg
with a h5T)ertonic solution in the presence of free oxygen,
whereby the poison was rendered innocuous by oxidation ;
or he withdrew the oxygen, and consequently suppressed
the oxidation-processes, by means of cyanide of potassium.
According to Loeb, it is the c3^olytic action in the cortical
layer of the egg due to the various acids mentioned — ^the
carbohydrates, the alkalis, the different extracts of organs,
and the lysin contained in the sperm of animals of the same
species — which accomits for the actual formation of the
membrane, and consequently for the abnormal development
of the organism, although imder certain circumstances the
segmentation of the egg may take place without membrane
formation. The merely mechanical factors also act, accord-
ing to Loeb, by means of cytolysis or cell-degeneration.
Formerly, Loeb objected to the definition of these factors
as stimuli, but recently he himself has classed them amongst
the formative stimuli.

At this point I may perhaps be allowed a digression.

Loeb, in the introduction and preface to his latest summary

on artificial parthenogenesis, writes as foUows : — " The fact

that the incitement of the development of the egg suggests

processes of ' stimulation ' is responsible for the devotion

of so much thought to this problem. I found out during

my researches on tropism and on the physiology of brain,

nerve, and muscle, that no decisive progress could be made

until the nature of the process of stimulation had been

elucidated. In spite of more than a century of research

and experiment we are still much in the dark, and this may

be ascribed to the fact that we cannot, at will, directly

observe what takes place in the nerve and its terminals

during stimulation. In the egg we can see the actual

processes at work. What is still more important, we can

confirm our conclusions by evidence derived from like

simultaneous experiments on a very great number of

13

194 THE MNEME

individuals. This suggested to me the idea that the study
of the artificial incitement to development might supply
the missing analogies for the successful analysis of the
processes in muscle, nerve, and other cells."

If now we ask the question whether Loeb by this method
succeeded in penetrating deeper into the nature of the
process of excitation, we shall have to reply in the negative,
even if we ignore, as we do, the purely hypothetical element
in his theoretical results. Apart from the statement that
we are able to observe directly with the eye the processes
of stimulation in the egg — which statement is much too
wide and, therefore, misleading — ^neither the discovery that
the formation of the membrane is, as a rule, the first visible
reaction of the initiated development, nor the equally
valuable demonstration that this membrane-formation is
induced by a superficial cytolysis of the egg, affords us any
real insight into the " nature " of this formative stimula-
tion, that is, into the chain of the concurrent chemical and
phj^ical processes. Further, we have to note that the
formation of the membrane which, according to Loeb, is
the first definitely essential process in the incitement to
development, may under certain circumstances not take
place, and yet the eggs may nevertheless develop. If
Loeb, in spite of his apparent failure to understand the
nature of the stimulation-process in this specially selected
instance, means to suggest the entire obliteration of the
words stimulation and excitation from the vocabulary of
physiology, we can only say that at the present stage of
our knowledge in the realm of muscle- lierve- and sense-
physiology such a procedure must lead to utter confusion.

Without doubt, excitation is at bottom a physico-
chemical process and nothing else, and the engram simply
a residual ph37sico-chemical modification. But as we are
still lacking any real insight into the physics and chemistry
of these processes, it would be the greatest mistake to discard
the serviceable definitions of stimulus-physiology and to
deceive ourselves about the distance which still separates
us from the goal of a purely physico-chemical interpretation.
For such self-delusion, followed by the inevitable isillusion,
would only serve the purposes of the vitalists.

THE INITIAL ONTOGENETIC ENGRAM 195

While in cases of experimentally induced parthenogenesis
we know the stimuli acting ecphoricaUy on the initial onto-
genetic engram, we are still in ignorance of the ecphoric
stimuli in those comparatively few cases where partheno-
genesis appears as a normal phenomenon, as, for instance,
with some of the Rotifera, Crustacea, and Insects, and
abnormally with the Starfishes. It may be that in some
cases the external stimulus is of a chemical or mechanical
nature, and that this acts during the passage of the eggs
from the abdominal fluid into the air or into the water.
But in other cases, as, for example, with the parthenogenetic
viviparous Aphides, the intervention of such an external
stimulus is impossible, as segmentation foUows immediately
upon the maturation of the ovum. Here, the probability
is that the processes of maturation themselves act ecphoric-
aUy on the initial ontogenetic engram immediately following.

In the asexual development of ferns generated from spores,
the moistening of the spores acts ecphoricaUy on the initial
ontogenetic engram.

When the initial ontogenetic engram is ecphorised by a
stimulus of any kind, the further course of the ontogenesis
proceeds in the main according to the principles enunciated
in Chapter VIII. The course, however, is affected by
external conditions, which play now an active, now a some-
what passive, part.

The temperature at the time of development plays a
passive but yet a very important part. It determines the
tempo of the entire metabolism, and consequently of the
plastic reactions ; on it depends the tempo of the entry
of the original excitations, and thereby, as we explained
on page 183, the tempo, but not the rhythm, of the entire
cycle of both original and mnemic excitations. By lowering
the temperature we are able to retard the ontogenetic course
in a most extraordinary way. Whether we can bring it
to a complete standstill without permanently injuring the
organism has been made doubtful by the recent investigations
of O. Schultze on eggs of Ranafusca. However that may
be, on the return to normal conditions — ^in this instance
normal temperature — ^the course of development progresses

196 THE MNEME

as usual, without requiring any fresh external stimulus to set
it going again. The amount of light, the nature of the
medium, and the supply of food play a similar, but usually
a less important, role in ontogenesis. Each of these factors,
however, may under certain circumstances acquire in some
specific phase of the ontogeny a greater importance, as when
the ecphory of certain engrams fails to take place along the
habitual path of successive association unless specific externcil
stimuli, acting as ecphoric stimuli, intervene.

For example, specific changes in the skin, giUs, and tails
of many newt larvae appear only when the young animal
is given a chance to get into direct touch with the atmo-
spheric air. By preventing this contact, by excluding the
larvae from the air by means of a wire netting placed beneath
the surface of the water, the modifications are inhibited,
but the animals survive as larvae, continue to grow, and
finally become sexually mature. The absence in such cases
of an ecphory dependent on specific external stimuH does
not, therefore, mean that the development is definitely
arrested, but only that certain portions of the mnemic
excitation-complexes remain quiescent until ecphorised by
stronger original stimuli.

A close investigation of these abnormal cases shows us
that this dependence of ecphory on external stimuli generally
possesses special biological significance in the way of service-
able adaptations. It is appropriate for the axolotl or the
triton to lose its gills and to be transformed into a terrestrial
animal when it has the opportunity of reaching dry soil.
The problem indicated is fascinating and important, but
its discussion would lead us somewhat far afield, as at the
moment we are not primarily concerned with the rise and
development of useful adaptations. It will be a separate
task to consider the Darwinian theory of natural selection
in the light of the general views elaborated in this work.
At this point I only wish to say that this separate exposition
will in no sense be an attack on a theory, which, in recent
times, has been so vehemently criticised. At most, some
few modifications will be suggested.

Reverting to those cases in which the ecphory of the

THE INITIAL ONTOGENETIC ENGRAM 197

morphogeneous engrams does not follow the normal course
described on page 184, but demands the action of an external
stimulus, we wish to point out that in such instances it is
a case of special adaptations, for in most cases we can prove
that the necessary external ecphoric stimulus is an integral
part of a previous engraphic stimulus which, in the corre-
sponding stages of development, has influenced the ancestors
of the organisms in question. Triton larvae usually lose
their gills, if they are allowed to come to the surface of
the water and so to gasp for air. It is not altogether
necessary for them to get on to dry soil and in that way
to expose the gills themselves to the fuller influence of the
atmospheric air. Speaking phylogenetically, the entire loss
of the gills and the transformations of the skin and the tail
must imdoubtedly have taken place imder a much more direct
and intense influence of the atmospheric air and of Ufe on
dry soil. Ontogenetically, a mere portion of these stimuli
now suffices to induce in the descendants of those primitive
tritons the ecphory of engrams originated by those stimuli.
Already it has been mentioned that the absence of ecphories
dependent on external stimuli does not imply the unquah-
fied arrest of the development, but simply the inaction
of certain portions of the excitation-complexes. We must
here ask how this is possible. Does not a simultaneous
engram-complex become ecphorised in its totality, or does
it still require, under certain circumstances, a specific
ecphory for its separate parts ? Referring to the general
discussion of ecphory on page 78, we note that the recur-
rence of a portion of an original excitation-complex does
not always involve the ecphory of the corresponding engram-
complex in its entirety (see diagram, p. 139). In some
instances one or other of the engrams of a complex cannot
become ecphorised at all along the path of simultaneous
or successive association, but only by the recurrence of
the corresponding original excitation. The stimulus Uber-
ating this recurrence may be an exceedingly weak one.
It may be, for example, in regard to a long-lost friend, that
neither a description of his person nor a reminder of situations
and events in which he played a part suffices to ecphorise

198

THE MNEME

his featiires in our memory ; but a slight pencil sketch of
his face may bring him before us at once. So as in the
case of many salamandrina a casual and short contact with
the atmospheric air ecphorises an excitation-complex which
manifests itself in such reactions as the resorption of the
gills, and alterations in the skin and tail. In the larvae
of many genera, if all contact with the atmospheric air is
prevented, the ecphory does not occur.

It may be well, therefore, in order to cover certain ex-
ceptional cases, to make the following additions to our
diagram on page 184 of the cycles of the normal morpho-
genesis (Diagram III) : —

III.

Phaser.

Phase 1.

r (or) + r (mn) are not capable

by themselves to ecphorise

in this special case the com-

plete engram-complex s (engr)

along the path of the succes-

sive association. The ec-

phory of a certain fraction,

that is, p" (engr) of this

mnemic excitation-complex

complex, requires the as-

.. /„.., ..^\ ^.fc.

sistance of an original ex-

^/

*

citation p" (or), generated
by an external stimulus.
Then r (or) -1- r (mn) + p» (or)
together exphorise s (engr),

P" (or)

originating the mnemic ex-
citation-complex s (mn)

plastic reactions :

plastic reactions :

morphological state r (z)

morphological state s («)

i

i

original excitation-complex

ori^nal excitation-complex

« ?„-\

s(or)

r \pi)

homophony between r (mn)
and r (or)

[homophony between s(mn) and
s (or)]

Phase, r

Phase. S

Jlorphological si-ate.
r'~*(Z), product cf
the. plash'c reactions
def-e.rniine.d by Me
mnamic exaraHon
com/j/ex r^'*(mjl)

above.: mne.m'ic

excii-af-ion
com a/ex r''*^n)
mid: original

excH'OTion
comple.xr''*M
below: original

liberal-edhyan
exi-ennal
stimulus

Engram
complex

(Er\<jr)

above: mnemic excifafion complex
Ae/ow ; original encltaHon complex

r'(z) ri(Z) ri(z) f*(z)

^

^51 (mn)
S' (mn)
53 (mn)

S* (mn)
7^5« (mn.)

SUZ) S2(Z) S^(Z) 5*(ZJS^a)

^ Si im.)
\ s* (<n)
' S£(cn)

200 THE MNEME

In the preceding diagram (IV) is illustrated more fully the
ecphory of an engram-complex s (engr) — ^for convenience
and simplicity limited to s' (engr) s* (engr) s^ (engr)
s* (engr) and s* (engr) — by the serial morphogeneous
excitations r '"*, and by the excitation c', which is liberated
by an external stimulus. For detailed explanation the
reader is referred to page 96, where is discussed the intimate
connection between the single components of successive
engram-complexes, and to page 140, where certain ecphoric
peculiarities are dealt with. Diagram II (p. 185) should
also be consulted.

A superficied study of the latter diagram might perhaps
raise the suspicion that the investigations in this book,
as far as their application to ontogeny is concerned, had
led to nothing more novel than a restatement of processes
already well known. Some may think that the processes
of evolution directly observable have been understood quite
as profoundly and analysed quite as keenly by previous
workers, without the results being obscured or confused by
mere theoretical formularising. Of theories which merely
paraphrase we already possess an abundance. A close
study of the diagram in question will, I think, immediately
disabuse the mind of this suspicion.

The introduction of mnemic terminology into the diagram
and the consequent inclusion of ontogenetic relations vmder
mnemic laws give these relationships a far wider meaning
than those which are based On ordinary original stimula-
tions possess. As the diagrams show, the ontogenetic
connections are made — at any rate at those points where
they cross — by the ecphory of engrams and engram-com-
plexes. But these ecphories, compared with the ordinary
original stimulations, have this peculiarity, that a complete
engram-complex may be ecphorised by a mere portion of
the excitation-complex. For instance, if in our diagram
IV, on page 199, the sign c' (or) signifies in newt larvae
an, original excitation generated by contact with external
air, in certain forms the engram-component s' (engr) may
be ecphorised without the appearance of c^ (or), either
by simultaneous association through the ecphory of s'"*

THE INITIAL ONTOGENETIC ENGRAM 201

(engr), or, later, by other ecphoric influences. In some
cases, however, ecphory depends entirely on the entry of
the original excitation c' (or). In the case of Salamandra,
the ecphory of the engram s* (engr), though often somewhat
belated, nearly always takes place without the appearance
of the original excitation c* (or). But with Siredon, the
ecphory is equally certain to remciin latent imtil e' (or)
appears.

The diagram, it is true, shows but one series of relations
during a specific ontogenetic process, but that is sufficient
for the realisation of what the process means. The process
may run out and reach the same goal, even if several com-
ponents have remained inactive. The reader is referred
to page 78, where the various possibilities of lens-develop-
ment are discussed ; especially is he asked to note those
numerous cases where lens-formation takes place even on
the eUmination of that stimulus of contact which has been
supposed to be the sole determining factor. In any con-
sideration of these ontogenetic processes, we must, there-
fore, take into account the many possibilities of relationship
which permit of collateral ways leading to the same end.
The mnemic conception of ontogenesis allows us to do this
in the fuUest way possible.

Another equally important advantage is derived from
the application of the mnemic principle to the connections
between the ontogenetic cycles. We see the entire onto-
genetic course imfold itself, in the presence and by the
power of mnemic homophony as a necessary implication of
the connections shown in the four diagrams, I-IV, on pages
184, 185, 198, 199. The idea of homophony offers a
path by which we may penetrate more deeply than has
hitherto been possible into the meaning of the somewhat
bewildering phenomena of regeneration and regulation.

I do not claim to have solved the deeper riddles of biology
These still remain. I think, however that I am justified
in assuming that in this book I have offered something
more than a mere paraphrase of the problems discussed
therein.

CHAPTER X

MORPHOGENEOUS MNEMIC EXCITATIONS IN
THE FULLY DEVELOPED ORGANISM

At the outset it would be well to make quite clear what
we mean by " fuUy developed " organism. Is an in-
dividual to be regarded as developed when it has reached
sexual maturity, or only when it has attained its full
growth ? The latter criterion is obviously inapplicable
to many forms of plant life where the growth is practically
unlimited. Further, in the case of many other organisms
of limited growth it would be diflftcult to apply this test.
For in many organisms a growth in thickness proceeds
for years after the arrest of the growth in height, and
ceases only when the other organs have long since passed
their highest point of development. In man, for example,
the cessation of vertical growth is followed by an increase
in volume of the bones, muscles, and other organs. The
criterion of the production of mature germ-ceUs is equally
impracticable, so far as the indication , of full organic
development is concerned. We need but think of those
numerous cases of so-called paedogenesis, where morpho-
logically undeveloped forms such as the larvae of Ceci-
damyia reach sexual maturity and produce young, in
order to be convinced that the criterion of sexual maturity
cannot be regarded as absolute.

Embryologists usually consider an organism fully
developed when most of its organs have so far matured
that any changes arising indicate simply a mere increase
in volume and not a further differentiation of function.
This definition is useful for descriptive purposes. But

MORPHOGENEOUS MNEMIC EXCITATIONS 203

phrases like " most of its organs " and " mere increase
in volume without further differentiation " express clearly
the conventional character of this division of the Ufe-
cycle of an organism. The definition, hke all those
based on separate characteristics, such as the attainment
of sexual maturity, of full growth, etc., is inadequate.

As long as the individual lives, physiological and
morphological changes continually appear, and although
we may be justified in summarily dividing the life-cycle
into two parts, one in which the course of the morpho-
logical changes is intense and rapid, and the other wherein
the tempo is remarkably retarded, we have to remember
that there exists no sharp boundary between the two
divisions. The transition from the one to the other is
gradual and almost imperceptible.

If we examine the course of movement in this transition
period, by reference to the diagram on page 184, we find
that the only real difference is in the increased duration of
the separate phases. The tempo of the plastic reactions
within each phase is more rapid than the transition from
one phase to another, the ecphory of new engram-com-
plexes. The length of the time-periods gradually increases.
After reaching a morphological state t (z), the mnemic
excitation t (mn) as well as the original excitation t (or)
remain homophonically active, imtil at last the ecphory
of the new morphogeneous engram (u) takes place. In
organisms of Hmited growth a time ultimately arrives
when a fresh ecphory of new morphogeneous engrams
is no longer possible, for the last hnk in the succession
of inherited morphogeneous engrams has been ecphorised.
We may call this engram w (engr). Its ecphory rouses
mnemic excitation w (mn), the plastic reactions of which
estabhsh the morphological state w {z). In this state
and in the original excitation w (or) determined by it,
the organism will remain until it dies. The organism is
now morphogeneously stationary in the phase w, within
which it passes only through cyclical changes.

A question here arises. At the beginning of phase w,
immediately after the morphological state w {z) has been

204 THE MNEME

reached, the ordinary homophony holds good between
the mnemic excitation w (mn) and the original excitation
w (or). The latter excitation lasts, of course, as long
as the morphological state w {z) lasts. Can the same
thing be said of the mnemic excitation w (mn) ? It
might be assumed that when the entire morphogeneous
succession of mnemic excitations had been completed,
the last link would gradually disappear. But the assump-
tion is a fallacious one, for the presence of the original
excitation w (or) must act ecphorically on the mnemic
excitation w (mn). The surest evidence for the perma-
nence of the homophony H ^ ;°5\ is adduced by the
"^ -^ w (mn)

continuance of reactions peculiar to this homophony.
In perfect homophony these reactions naturally do not
appear, but in imperfect homophony they are fairly
constant, though with varsdng results according to the
species. These reactions, now so famiUar to us, either
remove the incongruity or diminish it. If on the es-
tablishment of homophony it is a case of the re-building
up of parts which have been experimentally removed
or otherwise lost, the reactions are usually described
as regeneration. Many biologists apply the term " re-
generation " only to fully developed organisms. Corre-
sponding phenomena in earlier developmental stages are
labelled with other terms. The point is that they regard
as fully developed those states where the tempo of the
ontogenetic development has indeed become very much
retarded, but where the succession of morphogeneous
engrams has not yet been finally completed. The ex-
pression " fully developed state " is in this connection
used in a very loose and arbitrary manner.

For this reason, and also because no real necessity
seems to exist for distinguishing the processes of regenera-
tion in the different ontogenetic phases, we refrain from
giving a separate definition, and content ourselves by
describing all these reactions as regeneration.

In the next chapter we shall deal more fully with the
restriction of the generative capacity in the latter stages

MORPHOGENEOUS MNEMIC EXCITATIONS 305

of life. For the moment it suffices to point out that
the undoubted presence of the generative capacity in
the latest stages of existence must, in spite of all restric-
tions, be regarded as a reaction proving the efficacy of the
morphogeneous mnemic homophony, and, consequently,
the existence of a morphogeneous mnemic excitation in
all the stages of life.

We mentioned a state which one might feel justified
in terming " fully developed," in so far as no new ecphory
of inherited morphogeneous engrams occurs after it has
been reached. But this state does not involve the
cessation of every morphogeneous course. The course
which then begins, or, as it may be, continues, is a cycUc
one with regularly recurring phases, which manifest a
certain dependence on the periodicity of sidereal pheno-
mena, diurnal recurrences, and seasonal changes, and
even, as in the case of the Palolo worm, on the phases of
the moon. The periodic maturation of male and female
germ-products, which is generally accompanied by mor-
phogenetic processes manifested in the appearance of
various secondary sexual characters, belongs to this
category. Plants also show us similar cyclic morpho-
genetic processes in connection with the diurnal period,
and especially with the annual period. That these
cycles are not exclusively induced and regulated by the
periodic change of the external conditions, but that an
equal share in these adjustments belongs to mnemic
processes, may easily be proved by the fact that the
organic periodicity continues for some time after the
periodic change of the external conditions has been
eliminated by the cultivation of the respective organisms
under artificial conditions.

On page 54 we show how the genesis of " chrono-
geneous " ecphory must be conceived. Under normal
circumstances the ecphoric action of the periodically
changing external conditions is added to this chrono-
geneous ecphory. In certain plants, difficult or impossible
to force, the latter factor acts far more powerfully than
the former, but in others the case is reversed.

2o6 THE MNEME

Further exposition is here unnecessary. Still, it has
been useful to refer again briefly to those cyclic processes
which frequently warp ontogeny proper and survive it in
organisms of limited growth, in regard to which it might
be perhaps permissible to^speak of a " fully developed "
state.

CHAPTER XI
ENGRAM LOCALISATION AND REGENERATION

Dealing with the localisation of inherited engrams,
(p. ii6), we concluded that the entire inherited engram-
stock of each individuaUty-phase, whether this is initiated
sexually or parthenogenetically, is to be found within
the limits of a cell. It is most probable that the engram-
stock may be encompassed by something smaller than
the cell or even the nucleus of the cell. We called the
most minute morphological unit encircUng it a mnemic
protomer, but made no attempt at any precise morpho-
logical deUneation of this unit. We also found that in the
later course of an individuahty-phase, when the plant or
animal has become multi-ceUular, fragments cut from
any part of the organism seem in many cases to be in
possession of the entire inherited engram-stock.

But at the beginning of Chapter V we saw that the
regenerative faculty, even in forms especially capable of
regeneration, such as Planaria and Hydra, diminishes
towards the end of the ontogenesis as compared with
its strength at the beginning. In Planaria this diminu-
tion of regenerative power is evinced by the fact that
portions from the extremities of either front or sides,
devoid of nervous structiure, cannot regenerate the entire
individual, although sections from the rest of the body
may do so. In Hydra, any section of more than
J mm. diameter may regenerate the entire Hydra, but
sections of the tentacles, however large, lack regenerative
power. The restriction of the regenerative faculty ob-
served in these two cases is very sUght, but in other

207

2o8 THE MNEME

forms it is more pronounced. Many of the higher animals
such as the warm-blooded Vertebrata, possess, even
when fully developed, only a very limited capacity for
the regeneration of entire organs, although regeneration
of tissue may still take place in them to a great extent.
This raises the question whether this restrictive power
of regeneration and regulation is based on any specific
change in the localisation of the mnemic properties of the
organism, or whether it can be traced to changes which
occur during the ontogenesis.

The discussion of the problem of the localisation of the
inherited Mneme is reserved for a sequel to this book,
where it will be treated on broad general lines. Here it
is only necessary to show that facts, even those repre-
sentative of the most striking restrictions of the capacity
for regeneration and regulation during ontogenesis, do
not clash with our conception that each living cell, nay,
each mnemic protomer, from whatever part of the develop-
ing or fully developed organism it may have been taken,
is in possession of the entire inherited engram-stock.

We may at once rule out those cases of restricted
regeneration and regulation, where the regeneration is
tardily effected by part of an organism, but where, never-
theless, regeneration does occur. In regard to those
cases, there can be no doubt that the fragment of the
organism has been in possession of the whole body of
inherited engrams ; otherwise, it would not have been
able to replace the entire organism with all its morpho-
logical and physiological properties. Very belated regula-
tions and regenerations occur frequently with fragments
of eggs or in the early developmental stages of Echinoderms
and Amphibia, while with corresponding fragments of
Medusae, Amphioxus, or Teleostei, regulative adjustment
is immediate. In this connection should be mentioned
cases of regeneration in fully developed animals, such
as those observed by Przibram among the Crustacea —
Portunus, Porcellana, Galathea, etc. — ^where an imperfectly
regenerated third maxilliped was replaced after each
shedding by an increasingly improved new formation.

ENGRAM LOCALISATION 209

These cases indisputably prove the presence of the entire
inherited engram-stock in the parts capable of regulation.

They are of special interest in view of the problems
with which at the moment we are concerned, in so far
as they show how, even in the presence of the whole
body of inherited engrams, interference and disturbances
may so affect the morphogenetic adjusting processes that
the regeneration and regulation become perfected only
after great delay. If by reason of still greater obstacles,
regulation or regeneration does not follow, it would be
a hasty inference to say that this was due to mnemic
deficiency

If we define the processes of regeneration and regulation
as the sum total of plastic reactions which at homophony
effect the removal of incongruity, then the entire absence,
or the imperfection, of the regenerative process may as
probably be caused by a retardation or incapacitation
of the plastic reactions as by the absence of homophony
through a deficiency of the corresponding engram-com-
plexes.

Take the case of a pair of Weaver birds in captivity,
who give no evidence of their weaving capacity. It is
clear that we cannot infer the extinction or congenital
absence of this capacity unless we are sure that the birds
have aU the needful material for nest-weaving at their
disposal, and that the necessary preliminary conditions,
such as sufficient feeding, absence of disturbance, and suit-
able space, have also been fulfilled.

Absence of regeneration and regulation implies in
most cases simply the absence or insufficiency of definite
material for the neoplastic activity of the organism.
In other cases, the decrease of the regenerative capacity
coincides with the decrease of the neoplastic capacities
of the organism in general, that is, with a decrease of
plastic capacity in which mnemic processes do not come
into play at aU. Neither do these cases indicate a deficiency
of inherited engrams, or, as we may put it, they cannot be
quoted as furnishing evidence for a localisation of the
inherited engram-stock. After examining them, we shall

14

210 THE MNEME

have to see whether anything at all remains that cannot
be explained without the assumption of such localisation.
As we have seen, the absence or insufficiency of definite
building material, which in certain circumstances can
prevent regeneration or regulation, makes itself chiefly
felt in the initial stages of the various individuality phases.
For we note that the material for its life journey which
an organism receives from its parent is most frequently
so well packed and arranged that everything is close at
han,d for the later processes of differentiation. A master-
builder would never dream of mixing up all his materials
indiscriminately before he began to build. Boveri acutely
observes that " the development is siinplified, if already
in the ovum the different substances are so allocated
that each primitive organ directly receives just those
materials which are most serviceable for the process of
its further differentiation." Boveri discovered in the
ovum of the sea-urchin Strongylocentrotus lividus a directly
visible stratification of the egg-plasm in at least three
separate zones. Granted normal development, one of
them serves for the formation of the mesenchyma, the
second for the formation of the archenteron, the other
becomes the ectoderm. An equally distinct allocation of
material can be discerned in the eggs of Annelida, such
as Nereis, Myzostoma, and Lanice ; and of MoUusca,
such as Dentalium, Patella, and Ilyanassa. In them we
can see, in normal development, different material used
definitely for plastic reactions and the like. The removal
of certain parts is, at the early stages of development,
often equivalent to the elimination of specific material
which is present nowhere else in the organism and which
is essential for bodily growth. We need not be surprised,
therefore, that such a removal either makes definite
plastic reactions impossible, or delays them until the
metabolic processes of the organism have reproduced
the necessary material. In many instances, however, it
is only the fully developed maternal organism that is
capable of the reproduction of this material. For example,
in the snail Ilyanassa, the formation of the mesoderm,

ENGRAM LOCALISATION 2tx

according to Crampton, depends on the so-called vitelline
lobe. If one separates this lobe from the egg- before
segmentation begins or soon after, the mesoderm is not
formed. In all these cases we see not so much a locali-
sation of dispositions as a localisation of the necessary
material for these dispositions — ^Wilson's " cytoplasmic
localisation." It is not a case of the localisation of the
directing forces, that is, of the inherited engram-stock. We
are supported in this view by H. Driesch, who, in respect
of a special case occurring in Echinoidea, writes : " When
animals at their eighth and sixteenth cleavage stages
fail to gastrulate — for Boveri's researches permit us to
state this — this apparently results from a certain lack in
their material means, and not, as I previously assumed,
from some deficiency in their regulating capacity."

It is not at all difficult in a great number of cases,
described by some as " the restriction of the prospective
potentiality," to trace the decrease of the regenerating
and regulating capacity of the parts to a localisation of
the material used in the building up of the body. In
many cases this localisation, which may consist of a
specific plasmatic distribution or stratification, cannot be
ocularly demonstrated, but this does not at all justify
us in denying its actuaUty, and in ascribing the decrease
of the capacity for regulation simply to a localisation
of the inherited engrams. We are inchned to agree with
Boveri, when he says : — " We can with assurance say that
this stratification exists in all echinoderm-eggs of the same
cleavage type, although we are not able to see a trace of
it. This naturally suggests the idea that in the egg of
Strongylocentrotus, a still finer stratification exists than
those three zones which we are able to distinguish."

That this appHes, mutatis mutandis, also to the Cteno-
phora seems to me indisputable, if we are to judge by the
experimental data already to hand, although the ocular
evidence of such stratification has not yet been given.
A happy chance may reveal it to us some day. In any
case, even if the locahsation of what we have called the
building material is questioned, no reason exists for

212 THE MNEME

assuming that the lack of the regulating capacity in these
forms points to a regional allocation of the inherited
ehgiam-stock.

" We have already noted that the majority of those
cases where defective or insufficient building material is
undoubtedly responsible for the lack of the power of
regeneration or regulation occur mostly in the very early
stages of ontogenesis. But similar cases are found during
the intermediate stages of development, and also among
fully matured organisms, where one expects not merely
the regeneration of one part or organ by the whole, but
the regeneration of the whole creature by a part. In
the case of Hydra, it has been found that a dissevered
tentacle is unable to regenerate the entire animal. Here
the case may be one of the lack in the tentacle of certain
constructive materials essential for the building up of
the entire animal. The isolated tentacle, unfitted to
absorb nutrition, is unable to produce the material neces-
sary for regeneration. But in the great majority of
relevant cases it is only a question of the regeneration
of a small part or a single organ (limb, eye, etc.), by
an organism otherwise intact. In those cases the failure
to regenerate can hardly be traced to the lack of definite
building material. For it is hardly conceivable that an
organ, otherwise uninjured and with its functions in
perfect working order, should not be able to produce
the material necessary for the purpose of regeneration.

The problem then may be stated thus. In all Metazoa,
the regenerative capacity, varying of course according
to the species, undoubtedly decreases with the advancing
age of the individual. This decrease cannot be traced
to the lack of definite building material. Are we justified
in inferring from this mere decrease of regenerative power
a regional allocation of the inherited engram-stock, an
allocation which becomes increasingly specific as onto-
genesis advances ? I think we can answer this question
decidedly in the negative, even although we are not yet
in a position to enumerate all the factors which make
during the course of the individual life for the diminution

ENGRAM LOCALISATION 213

of the regenerating capacity. For the decrease of the
regenerating capacity of a specific organ is not in any
definite relation either to the development of the organ
or to the differentiation of its tissues. The researches
by Barfiulh have shown us that one tadpole was able
to regenerate only extremities — climbs and tail — ^that had
just begun to emerge, while other individuals oiE the
same species with greater regenerative power were able
to replace more highly differentiated extremities. Spal-
lanzani found that even young frogs and toads were
capable of re-growing limbs that had been removed. But
these were exceptional cases. It is clear that the poWer
of regenerating the extremities is exhausted in individuals
of the same species at different times, and that this
follows without any constant regard to the differentiation
of the tissues in the organ specifically concerned. A
trout-embryo is able before absorption of the vitelline
sac to regenerate the aheady well-developed and differ-
entiated tail with anus and the so-called urethra ; but
a Uttle later, when the vitelline sac has been absorbed,
it is no longer capable of doing so.

With the Anura, the power to regenerate extremities
ceases in the fully developed state ; but with the Urodela,
this capacity is retained by the sexually mature and fully
grown animal. Among the more highly differentiated
forms of the Salamandrina we meet with a considerable
restriction of this capacity. Formerly, it was thought
that in the fully developed Salamandrina perspecillata, the
power of regeneration was entirely absent. But this is
not so. Kammerer discovered (Z. /. Ph., vol. xix, 1905) »
that the larvae and young ones of this species required
a much longer time for the regeneration of the legs and
tails than all the other Urodele larvae examined for
this purpose ; and that in the adult Salamandrina per-
spicillata the regeneration proceeded with extraordinary
slowness, but that regeneration did take place.

Taking the Urodela, whose regenerating capacity is
well developed, and comparing the behaviour of individuals
« Zentralblatt fur Physiologiet

214 THE MNEME

at different ages, we generally find a retardation of the
regenerative processes commensurate with the age. In
order to furnish reliable data, I made a number of ex-
periments, the detailed results of which need not here
be given. It is sufficient for our present purposes to note
that larvae of Triton alpestris, 25 to 32 mm. in length,
and in possession of fully developed extremities, required
on the average four weeks for the regeneration of an arm
with the distinct formation of all the four fingers ; while
fully developed animals, kept under the same conditions,
required on the average seven and a half weeks to reach
the same stage. To re-establish the normal relativity
of size took five weeks with the larvae, and four to five
months with the fully developed animals. Kammerer in
his recently published essay sets out similar results.
The essay is of special value to us because it compre-
hensively deals with the dependence of the regenerating
capacity of Amphibia larvas on age, developmental stage,
and specific size. He also notes that " the Urodela
regenerate bodily parts capable of regeneration more
slowly in the adult state than in the larval stage."

We thus have a factor retarding the processes of re-
generation, whose power increases with advancing age.
We shall not at the moment examine the nature of this
factor, but content ourselves with the assertion that it
has nothing to do with a mnemic defect. But we may
ask whether this factor, which so often retards the process
of regeneration, is identical with that which in other
forms counteracts it entirely and causes it to disappear ?
Restricting our observations to the group of Siredon,
Triton, Salamandrina, and Anura, the question may
without hesitation be answered in the affirmative. And
if, on further survey, we meet with similar instances in
other animal groups, we shall not hesitate to draw the
same conclusion, without thereby committing ourselves
to the position that the obstruction and entire inhibition
of regeneration must always and under all conditions be
traced to one and the same fundamental cause.

Jt is not our present tagk to di3cover the causes of

ENGRAM LOCALISATION 215

the decrease of the regenerative capacity during the
individual hfe. We have only to prove that the decrease
does not depend on a regional distribution of the inherited
engrams. But I should like to say that, in my view, the
decrease of the power of regeneration does seem to rest
partly on a gradual decrease of the capacity to fulfil the
larger neoplastic functions, as may be observed in the
course of the individual life, particularly in organisms of
Hmited growth. The more highly organised and the
more complex the organism, the more this becomes
apparent. It reaches its extreme manifestation in the
remarkable fact that in the genus Homo the female
becomes incapable, towards the middle of her normal
individual life, of fulfilUng any longer one of the cardinal
functions of the organism, namely, the production of
germ-ceUs. In Hke manner pathological neoplastic re-
actions, such as tumours, are impeded concomitantly
with advancing age until they finally cease altogether.
It is well known that a senile carcinoma grows very slowly.

It may be that some will point to the fact that the
presence of a central nervous system and the undisturbed
transmission of nervous energy are necessary to certain
regenerative processes, and thereby infer, since certain
inherited engrams are deposited in the central nervous
system during the course of ontogenesis, that on the re-
moval of these organs the engrams also disappear, and with
them the possibihty of specific processes of regeneration.

This position is based on a false conception of the
place the central nervous system occupies in relation
to the plastic reactions of the other tissues. In the fixst
place, experiments bearing on this problem seem to have
definitely estabUshed that in the early developmental
stages all the organs develop independently of the already
existing central nervous system, and that they display,
where necessary, regenerative capacity. The reader is
referred to the experiments by Loeb, RafEaele, Harrison,
Barfurth, and Rubin, and to the interesting work on
anencephalous amyelous frog larvae, by Schaper, which
is confirmed by Goldstein. Even the muscular system

2i6 THE MNEME

develops and regenerates in these early stages indepen-
dently of the central nervous S3retem and the spinal
ganglia, as experiments by Harrison, Schaper, and Gold-
stein demonstrate. In the later stages of development
certain modifications are effected, as Rubin {Archiv. f.
EntwicUungsmSch, vol. xvi, p. 71, 1903), clearly has shown.
He points out that " the elimination of the nervous
system in Siredon pisciformis does not prevent the timely
start and the first stage of regeneration," but that, " later,
the defective innervation, or the fact of the missing
function, expresses itself in an increasing retardation of
regenerative processes until ultimately the power to
regenerate ceases." We must add to this that the arrest
of regeneration in later stages, on the elimination of
nervous influence, expresses itself not simply in the size
of the regenerated limbs and organs, but also in the
fact that the various tissues are affected in very different
degrees. Rubin discovered that " the elimination of
the influence of the nervous system was most apparent
in the muscular system. Here regeneration stopped
completely between the tenth and the twelfth day, before
the formation of specific muscular substance had even
begun." In order to appreciate the significance of this,
we have to ask how the muscular apparatus, on elimination
of all nervous influence, behaves in an organ which is
not stimulated to regenerate by some further interfer-
ence ? The answer is that the muscular apparatus becomes
atrophied. Whether this is merely an atrophy of muscular
power — a view which might be very well defended in
spite of conflicting arguments — or whether in the atrophy
the lack of the " trophic " action of the nervous system
also plays a part, is a question the answer to which would
lead us too far away from our main argument, and, as
far as the present problems are concerned, is of no funda-
mental importance. The fact itself, however, suffices
to explain why a paralysed muscular apparatus is in-
capacitated from regeneration. It is simply becafuse its
energetic condition has been radically changed by the
severance of its nerves, and an abnormal state created.

ENGRAM LOCALISATION 217

If anything is remarkable here, it is surely not the
cessation of regeneration, but the fact that, as Rubin
notes, regeneration can be observed at all in the initial
stages in such a muscular apparatus.

According to Rubin's investigations, in the other-
tissues of a regeneration-stump, where the nervous in-
fluence was eUminated by the cutting through of the nerves,
regenerative phemonena ceased, although relatively later
than in the muscular apparatus. The growth of the skin,
however, continued until the whole regeneration-cone
was covered, but further growth in thickness did not
take place. The growth of the connective tissue was
but slight. A very much restricted regeneration of the
cartilage commenced when regeneration of the muscular
apparatus had already ceased. In the vascular system,
the capillary vessels formed during the first ten days
were enlarged and became filled with blood. There was
no further formation of vascular capillaries.

In this case, that which gradually impaired the regen-
erative power of the remaining tissues was not the absence
of the nervous influence, but most probably the general
disturbance of the energetic condition caused by the
cessation of muscular regeneration.

The impossibility of muscular regeneration precludes
at the outset the removal of the incongruity in the homo-
phony, which in these cases would occur only if the
compUcated regenerative processes could be carried out
by the harmonious co-operation of all parts. The reactions
tending towards the removal of the incongruity cease on
account of the incompleteness of the co-operation. That
the absence of nervous influence does not seriously affect
the regeneration of slight defects in this or that tissue
is proved by numerous observations and experiments.
Wounds of various kinds heal as well in Umbs that are
deprived of nervous influence as in those with nerves
intact. In bone fractures, for example, an equally
perfect regeneration takes place in paralysed as in
non-paralysed extremities.

I may briefly mention the well-known experiments

2i8 THE MNEME

by Herbst on Crustacea, which demonstrated that re-
generation of the eye takes place only when the optic
ganglia of the amputated eye are allowed to remain.
It appears to me that more has been read into these
highly interesting experiments than is actually contained
in them. Embryology teaches us that, at the develop-
ment of the paired, composite eyes of the Crustacea, the
eye proper and the ganglion opticum emanate from a
common ectodermal growth. It follows that to extirpate
the coherent whole, both the eye and the ganglion opticum
must be removed. It is quite unimportant that this
latter part of the optic organ establishes in some forms
by secondary processes of growth a somewhat separate
position, thus presenting itself to the untrained observer
as something independent. It is now one of the best
known data of regeneration that an organ is regenerated
the more readily if parts of it are left behind in the
organism, than if the whole is radically removed. In
the latter case, in forms with only a moderate power
of regeneration, no regeneration whatever takes place.
Thus, according to Philippeaux, the Urodela, whose
regenerative capacity is comparatively great, regenerate
their extremities only if parts at least of the scapula or
of the pelvis have been allowed to remain in the body.
Minor defects in the eye may be remedied by regeneration,
but the Urodela are unable to regenerate the whole ball.
Throughout, the general rule holds good that the regenera-
tive capacity, especially of the higher animals, becomes
more and more restricted as their life proceeds. That
the Crustacea are able to regenerate even a part of the
optic organ is an imposing performance compared with
those possible to other highly organised animals. That
they cannot regenerate the radically extirpated eye, that
is, the eye with the ganglion opticum ontogenetically
belonging to it, only proves that their regenerating
capacity has its limits ; it does not prove that the ganglion,
which is, of course, connected with the central nervous
system, exerts any specific formative influence. This is
especially important, as during these experiments the

ENGRAM LOCALISATION 219

central nervous system proper — ^the brain — was left
intact and in function. Herbst has proved that a great
number of Crustacea can regenerate the organ if the
eye is not radically extirpated, that is, if the ganghon
opticum is allowed to remain ; but that they never do
so if both eye and ganglion opticum are removed. In
this case of incapacity to regenerate the eye, Herbst
noted that an antenn£e-like organ sometimes sprouted
from the seat of amputation.

Similar " heteromorphoses " have been not seldom
observed in Crustacea and Arthropoda. In such cases
a hmb may be replaced by one less differentiated. This
is t3?pical where the segment and the new formation are
in close proximity. In other cases, the replacing for-
mations are in their arrangement more intricately in-
volved. At some future time I shall try to show in
a continuation of the Mneme that these apparently
enigmatic phemonena are also open to explanation,
and that they conform to general laws. The discus-
sion of the subject here would lead us too far away
from our main purpose, but I should like to make clear
that, so far as these cases are concerned, I as Httle deny
a definite influence of the nervous system as in those
instances where the tempo of the regeneration is strikingly
retarded by the cutting of the nerves. The point is that
these cases do not prove that the nervous system exercises
any " specific formative " influence.

Przibram, in the second volume of his Experimental-
zoologie, gives an exhaustive account of the data rele-
vant to regeneration. It is interesting to note that on
page 224 he presents a conclusion which closely harmonises
with my own. Here are his words : — " If we keep in view
the fact that, on the one hand, regeneration is accom-
pUshed only by parts capable of growth, and that, on
the other hand, the regenerated parts develop out of
the growing tissues themselves, it is obvious that the
favourable influence of the nerves must be regarded as
simply stimulating growth, and not as a specific formative
/orce. Among the mammals, as is well known, the

220 THE MNEME

promotion of growth by the influence of the nerves is
a common occurrence. Also during the heterochely of
the Crustacea it occurs in a noteworthy manner."

In this present chapter I have tried to show first, that
the restrictions of the faculty of regeneration, which appear
in the course of ontogenesis, are not to be explained
by a regional allocation of the inherited engram-stock ;
and that, secondly, the facts of regeneration, from which
many authors infer a specific creative or formative in-
fluence by parts of the nervous system, may be interpreted
quite differently.

We feel, therefore, justified in assuming that each cell,
nay, each mnemic protomer, is in possession of the entire
inherited engram-stock of the respective organism. On
the other hand, in order to avoid misunderstanding,
I wish to reiterate that I regard the ecphory of each
inherited, as well as of each individually acquired,
engram as locahsed, that is, as bound up with specific
conditions of environment.

The reaction of pecking at grains and other small
objects, which newly hatched chickens manifest, must
be understood as evidence of the ecphory of an inherited
engram. But this ecphory is possible only where vision
is unimpaired. There must be at least one intact eye
with its nervous connections. The photic stimulus of
the grain liberates in the irritable substance of the retina
an excitation which, by conduction, releases in other
parts of the central nervous system another excitation
which acts ecphorically in its own specific area. But
this localisation of the ecphory — the existence of an area
proper of an ecphoric stimulation, whence it radiates
through the remainder of the organism — ^has nothing to
do with a regional allocation of the inherited engrams.
The predominating position in ecphory of the " area
proper " is explained quite simply by the fact that the
state required for the ecphory of the energetic condition
is here first reaUsed, and this state cannot become realised
at all if the area proper is either missing or badly damaged.

CHAPTER XII

ENGRAM DICHOTOMY IN ONTOGENESIS

Here and there in our considerations (pp. iii, i66, i88)
we, have been confronted with the fact that the engram-
successions, which usually are characterised by an un-
broken, unihnear, and non-reversible arrangement, may
occasionally spUt into dichotomies. Engram dichotomies
may be divided into two classes, namely, those whose
branches can be ecphorised simultaneously, and those
whose branches can only be ecphorised alternatively.
The former are of frequent occurrence in ontogenetic
engram-successions (see p. 187, on the grouping of
engrams during egg-segmentation), but it is with the
latter in their relation to ontogenetic engram-lines that
we are just now chiefly concerned. The reader will
remember that we set out in diagrammatic form the
structure of an alternative dichotomy, taking as our
illustration the two versions of the line from the Rubdiydt
of Omar Khayyam —

Ao. a Noose of Light.
The Sultan's Turret^

\with a Shaft of Light.

In its essential points this diagram is valid for all
alternative dichotomies, including those that determine
plastic reactions. We refrain at this point from enquiring
into the modes of genesis of alternative dichotomies in
ontogenesis. We shall here content ourselves with the
examination of those cases in which the choice of alter-
native is prompted by external influences, and defer

222 THE MNEME

until the next chapter the analysis of those dichotomies
which are generated by cross-breeding.

Take first the case of the honey bee, Apis melUfica.
It is well known that in regard to the females there are
two possibilities in the development of the fertilised
bee-ovum. Either a sexually mature queen emerges,
or an imperfect female worker. These forms differ
strikingly from each other in their structure and in their
instincts. In the worker the genital organs are not
fully developed, but compensation is given in a number
of positive characteristics, among which are the gathering
apparatus, the pollen baskets and bristles on the hind
legs, and the wax pockets on the medial abdominal keel.
In the queen, however, we find, with a full development
of the genital organs, an absence of the characteristics
just mentioned. Further, there is a reduced development
of the trunk, of the masticatory organs, of the salivary
glands, and of the wings. The shape of the sting differs
from that of the worker. The queen is also deficient
in the instincts connected with the rearing of offspring,
the building, feeding, and food-gathering instincts. She
has become simply an egg-laying machine. That is,
the physical and dynamical qualities, which we find
united in the female solitary bee, have in the case of
social bees been apportioned to two forms of female, the
queen and the worker, sharply distinguished according
to the perfect or imperfect development of the genital
organs.

If now we note the ontogenesis of these two forms,
we find that each fertilised egg is capable, according to
circumstances, of producing either a queen or a worker.
It has long been known that a grub belonging to a queen-
cell, if fed on worker's, that is, unmasticated food, will
develop into a worker ; and vice versa, a worker-grub,
if fed on queen's food, will develop into a queen. The
point is that the kind of food determines the nature of
the reproductive system.

It takes from three to four days for the egg of the
honey-bee to hatch. On the eriiergence of the grub, it

ENGRAM DICHOTOMY IN ONTOGENESIS 223

is fed by the workers for six days. Then follows the
sealing of the cells by the nursing workers, the spinning
of a chrysaUs-case by the grub, a longer or shorter period
of rest and pupation, and finally, the emergence of the
imago. Klein has demonstrated experimentally, and
V. Buttel-Reepen has confirmed his conclusions that, if
a grub is fed during the first thirty-six hours of its develop-
ment with worker's food, and then afterwards with
queen's food (we shall refer later to the specific chemical
composition of the foods), the short and early period
of feeding on worker's food does not produce any marked
change, that is, there is no ecphory of the " worker-
engram," but the development is typical of that of a queen.
It turns out differently, however, if the time in which the
grub is fed with worker's food is increased by one or two
days. We then get a queen with unmistakable signs of
worker-characteristics. Finally, grubs which have been
four and a half days in workers' cells and fed during that
time on worker's food, being then exposed to different
conditions, and submitted to the influence of queen's
food for the short remaining time of the larval stage,
show marked characteristics of both worker and queen.

Klein took worker-grubs of twelve to thirty-six hours'
old, fed them for two days with queen's food, and for
the succeeding day or day and a half with worker's food,
with the result that he obtained workers sUghtly modified
by queen-characteristics.

These experiments show that the ecphory of either
of the two engram-branches is influenced by the external
stimuli of nutrition. The accompanying diagram may serve
to illustrate the case.

It is assumed in the diagram that the dichptomy com-
mencing in phase four of the respective engram-complexes
is of such a kind that the possibility of transition from
one branch to the other remains until phase 9, and is
only excluded from phase 10 onwards. Such an oscillation
of the ecphory between two branches or even the simul-
taneous ecphory of parts of both branches manifests
itself in mixed reactions. In the case of the honey-bee.

224 ^^^ MNEME

we obtain under such conditions " transition-animals,"
that is, queen-bees with characters of workers, or workers
with characters of queens. The reader is referred to
the writings of Klein and v. Buttel-Reepen, in which
the experiments on bees are fully described and illus-
trated. Similar mixed forms are possible in the case of
ants also, as presently we shall see.

Phaw lSS4Gt789 10 11 IS

m

/
/

/
/

k

A

d — e — f — g — h — i

a — h — c

\l I . . ,

\

\

Something should be said on the nature of the stimu-
lation, which determines the initiation of the alternative.
In the case of the honey-bee, the result depends entirely
on the quality of the food supplied. There can be no
doubt that the stimuli in question are chemical. Accord-
ing to Planta, the queen's food as compared with that of
the worker's is, on the average, 5 per cent, richer in
albumen and 7 per cent, richer in fat, but 11 per cent.

ENGRAM DICHOTOMY IN ONTOGENESIS 225

poorer in sugar substance. It is not difficult to imagine,
however, that these quantitative variations are not the
sole determining factor, but that an admixture of specific
ingredients, inclusive perhaps of certain excretive products
of the nursing-bees, represents the ecphoric stimulus for
this or that dichotomy of the engram.

In the case of the ant where, according to Janet and
Wheeler, a qualitative apportionment of the nourishment
is hardly possible, it is most probable that the dichotomy
depends entirely on the admixture of such specific ingre-
dients. Ants are similar to bees in this respect. Re-
garding them, we are not as yet in possession of such
complete experimental data as in the case of bees. We
have to rely more on inferences from comparative obser-
vations and experiments. These, however, are in perfect
harmony with the experiments obtained with bees.

In 1874, Forel, dealing with various species of ants,
described forms intermediate between females and workers.
Since then, Wasmann has given a sufficient explanation of
the appearance in overwhelming numbers in certain nests
of the SQ-called pseudogynes, which are intermediate forms.
He has proved by observations extending over many
years that in these cases the appearance of the pseudogynes
is casually connected with the presence of certain ant-
guests, the Lomechusa, Xenodus, and Atemeles. These
beetles, tolerated by both the Formica and the Myrmica,
decimate the eggs and larvae of their hosts. A sudden
and appreciable reduction in the development of the
worker generation takes place. The ants thereupon try
to remedy the deficiency by rearing into workers all the
remaining larvae of the immediately preceding generation
originally destined for queens. Viehmeyer was able to
confirm this fact by a control-experiment. He placed
in a non-infected nest a queen which, living in a colony
infested with Lomechusa, had produced pseudogynes for
four years. From that hour the pseudogynes disappeared
from the progeny of this female, only pure full females
and pure workers being produced.
There is one difficulty of which mention should be

15

226 THE MNEME

made. With bees and ants, as also with termites, these
cases always refer to that one branch of the dichotomy
which leads to the development of individuals restricted
as to their generative capacity. Are we to be allowed to
regard this branch as a succession of inherited engrams,
despite the fact that the individuals in which the engrams
are aroused into activity appear to be excluded from the
reproductive group of the species ? The answer would
have to be in the negative for all those cases where we
had to assume that the dimorphism or polymorphism had
become developed only after the sterility of the branch
had set in.

But Herbert Spencer has already made out a strong case
in favour of the assumption that the reversed temporal
relation has ruled for the main forkings of the branches.

The point is illustrated very clearly by reference to the
honey-bees, where, as already mentioned, most of the
distinctive anatomical characteristics of the queen and of
the workers are in the case of the more simply organised
humble-bees united in the one individual. The same
thing holds good in regard to the instincts. Buttel-
Reepen, referring to the difference in instinct between
the queen and worker forms, says that " the main change
takes place in the queen, who sinks down from her high
position as universal provider, loses all her peculiar
instincts, and becomes simply an egg-laying machine.
The workers, however, retain nearly all the instincts of
their previous female nature, such as the building, feeding
or gathering instincts, etc., and lose only the sexual
instinct, in place of which they gain what has been called
the instinct of attachment to the hive-mother, which
involves the ispecific nutrition of the latter." Of course,
it cannot be denied that, besides the retention of already
existing characteristics, new organic properties and in-
stincts have also been acquired in the presumably sterile
branch. But this offers no insuperable difficulty to our
conception, since the worker caste is by no means entirely
excluded from the reproductive circle of the species.
As e^l'ly as 1874 Fovel reported cases of egg-laying by

ENGRAM DICHOTOMY IN ONTOGENESIS 227

ant-workers, and his observations have since been con-
firmed by Lubbock, Wasmann, Viehmeyer, Tanner, Reich-
enbach, Wheeler, Miss Fielde, and others. Silvestri found
the same thing among the workers of the termites, and
observations by Grassi prove that the termite soldiers
lay eggs. Escherich, who has studied most closely the
biology of the termites, has no doubt that further positive
evidence of the kind indicated will accumulate as soon
as these problems are definitely attacked. Finally, we
have to note in regard to bees that among the workers
absolute sterility is found only in one species of honey-
bee. Apis mellifica, and even of this species it is said
that among the Egyptian variety {Apis mellifica fasciata)
egg-laying workers are frequently found with the queen
in the hive. Among other varieties of honey-bee, whose
workers do not normally produce eggs, the absence of
a queen and of a brood capable of reproduction means
that eggs which develop parthenogenetically are laid by
the workers. From all this it follows that among the
social insects the workers are not entirely excluded from
the reproductive circle, and that even in the extreme
case of the honey-bee, where the specialisation of the
worker seems complete, the exclusion, especially with
certain varieties, is by no means absolute.

Up to now we have considered alternative dichotomies
where both branches are equivalent, in that both under
the normal conditions of life reach ecphory either in
this or in that individual of the species in question. In
these cases the ecphory of this or that branch depends
simply on the presence or the absence of a definite original
stimulus.

Such dichotomies may be defined as being in equili-
brium, and may be contrasted with those where the
branches are not equivalent. In the latter case the
ecphory normally takes place almost exclusively in
the one succession branch, while in the other branch
ecphory is realised by but a few individuals of the.
species, and then only under special conditions, which
mav be furnished by human agency.

228 THE MNEME

The reader is referred to the example on page 168,
where the dichotomy of successions of individually acquired
engrams was explained by the example of a poem learned
by heart in two versions. When there is quite a distinct
break between the learning of the one version and the
other, and when for some reason or other we prefer to
use only the second version, this branch becomes in time
the main path into which the ecphory naturally turns
at the point of bifurcation. The other branch is still
there, but a specially strong impulse is required to force
the ecphory into its path. To these two succession-
branches of individually acquired engrams correspond, in
the case of inherited successions, the branch that was
ordinarily used by our remote ancestors, and the branch
that gradually was preferred by later generations. These
dichotomous branches may fitly be described as " ata-
vistic " and " recent." The former may be obsolete,
but the point is that it still exists.

A word on Atavism or Reversion may not be out of
place. It must at once be admitted that it is difficult,
even impossible in many cases, to decide whether this
or that ontogenetic abnormity — a word to be understood
in its widest meaning — should be regarded as atavistic
or not. The difficulty is especially great in cases where
the ancestral line of the form is not at all or insufficiently
known, or where we are entirely ignorant of that part
of the line supposed to have possessed the presumably
atavistic peculiarity as a normal characteristic. The
inference that this abnormity is of an atavistic nature,
and that consequently the unknown ancestor of the
form showing the abnormity must have possessed this
morphological character, is very often quite fallacious
and hardly ever demonstrable. Critical caution is fully
justified, but it betrays itself as prejudice when it re-
presents the idea itself as a mere notion, vulnerable,
problematical, and antiquated.

For an example, we may take the hornless cattle of
Galloway and Suffolk. Although they have been without
horns during the last hundred or hundred and fifty years,

ENGRAM DICHOTOMY IN ONTOGENESIS 229

they axe undoubtedly descended from a horned stock.
Occasionally calves are born with loosely attached horns.
The phenomenon may rightly be regarded as a case of
atavism. The cbnclusion is equally valid in the case of
the hornless South Down sheep, of whom mjde lambs
with small horns are occasionally born. Sometimes the
horns grow to their full size, but often they are simply
attached in a loose way to the skin, from which in some
cases they drop oft altogether.

Many other examples of undoubted atavism might be
given, but to be absolutely sure of our ground, it is essential
that the ancestors whose characteristics reappear in
this atavistic form should be known to have existed.
A merely hypothetical construction is insufficient, but
even where hypothesis is used, the case may be adequately
strengthened to the point of conviction by arguments
from comparative anatomy, from embryology, and from
general biology.

We regard the reappearance of horns in a stock of
cattle or sheep that has been hornless for generations as
an instance of the emergence of a character which for
some reason or other has died down, but not out. Surely
it is permissible to speak of such cases as illustrative
of the dichotomy of engram-successions.

In the accompanying diagram, k, I, m, n signify the
engram-complexes in their aspect as the sum-total of all
the inherited engrams extant during the respective phase,
minus the engrams connected with the horn development
during the critical period when it is a question of horns
or no horns. The engrams of horn development we
denote by the engram succession a, j8, y, S. We then
obtain the following diagram : —

{Atavistic Branch)
(/ + a) — (m +p) — {n + y) —o + S) —

/

h

\

m-

(Rtcent Branch)

23Q THE MNEWE

A strongly pronounced dichotomy of the engram-suc-
cessioijs can be seen to exist. If now a plastic or motor
acquisition is refashioned, moulded into something new
and not simply discarded, the diagram shows it in this
way :—

(/ + a) — (m + j3) — (« + y) — (o + S) —

k

(/ +Ox) - (« + j8,) — (« + y^) — (0 + 8,) —

A series of special cases may be regarded as definite
examples of the first-mentioned group, where we are
concerned with the reappearance of characters which in
the later generations have not usually attained develop-
ment. There are cases in which such an arrested develop-
ment affects not merely a few, but the great majority
of the components of the engram-complexes, so that at
a certain stage an almost complete arrest of development
takes place. Apart from mere increase in bulk and one
or two unimportant alterations, the organism remains
permanently at this stage, and becomes sexually mature
in this state. This phenomenon is called Neoteny. As
we possess in this group significant and well-authenticated
examples of atavism and of its experimental generation,
it may be well to deal with it a little more fully.

We have already noted (p. 130) that the female
and male of the Mexican salamander, Amblystoma
tigrinum, usually become sexually mature in their larval
(Axolotl) stage and remain permanently in this state,
but from observations made in 1865 by Dumeril in the
Paris Jardin d'Acclimatisation, we know that young
salamanders go on land occasionally, lose their external
gills, develop into true salamanders of the Genus Ambly-
stoma, and are able to propagate in this state. The facts
were fully elucidated by the thoroughgoing experimental
researches of Miss M. von Chauvin, on the strength of
which Weismann recognised the maturation in the axolotl-
stage as a case of typical neoteny. Interpreting this

ENGRAM DICHOTOMY IN ONTOGENESIS 231

neoteny in the light of our theory of engram-successions,
we may construct the following diagram: —

Fhue

1

2

e

8

t

B

»

Engram-complex

a —

6 —

c

d —

e —

/-

g —

Phase

8

9

10

U

18

IS

14

Engram-complex

A —

» —

A —

I —

m —

« —

—

Phase

Engram-complex

16

16

? —

17 18

Ji-

ll

?3

80

?4

Let the letters signify the sum-total of the ontogenetic
engram-complexes until the close of the transformation
into the land species in phase 16 ; then only the engram-
complexes 5',, q,, qj, q^, in which the changes are quite
unimportant, are successively associated with the engram-
complex q. In order that this entire engram Une may be
ecphorised and so manifest itself in corresponding plastic
and motor reactions, a special external impulse is required
in the course of phase 10 (p. 131). This may be
given in the lack of oxygen in the water inhabited by
the larvas, so that they are compelled to breathe not only
with their gills, but also with their lungs. Or if this
does not sufl&ce, the removal of the animals from the water
to damp moss or mud is usually effective. In the absence
of either of these external impulses, no ecphory of the
engram-complex I takes place at phase 10, and the state
symbolised in the engram-complex k becomes altogether
or approximately permanent. In the following diagram,
this permanency is indicated in the fact that the anno-
tations kj, Aj,, ky k^, etc., denote simply the significant

232 THE MNEME

changes which occur after the ecphory of the engram-
complex k.

In phase lo the diagrammatic indication of the existent

A-

dichotomy is given by kC

But according to Miss von Chauvin's investigations,
even when the animal in its development has entered
on one branch of the dichotomy, the ecphory may, on
the action of appropriate original stimuli, pass during
the phases li to 17 from the neotenic to the atavistic
branch. The process is indicated by the connecting lines
between / and A,, k^, k^, k^, kj, k^, kj. The comparative
length of the lines may be taken to indicate that in these
later stages the development passes from- the neotenic
to the atavistic branch, with difficulty increasing propor-
tionately with age. The possibility of the movement
from one line to the other ceases when the animals attain
sexual maturity, a point we assume for both branches
in phase 18, and marked by the affix +y. In the atavistic
line the index of the sexual maturity engram-complex
is ^2+1 >■ ill the neotenic branch the index is k^^^
According to Miss von Chauvin's observations, it would
be possible to connect engram-complex k^—kj with m,
k^—kj with n, and perhaps k^—kj with 0, just as I is
connected with k^—kj. For although the ecphory may
have gone some way along the atavistic branch, mani-
festing itself in a series of reactions, it is still possible in
the case of the amblystoma for it to cross over again to
the neotenic branch ; although the process grows more
and more difficult the further the ecphory has moved
from the engram-complex /.

The reader should refer to the extremely interesting
observations by Miss Marie von Chauvin in her original
work of 1885, and read them in conjunction with the
diagrams on page 224 of this book. In the atavistic
branch we assume that the transformation into the
terrestrial type is completed at phase 16, after the ecphory
of the engram-complex q and on the initiation of the

EN GRAM DICHOTOMY IN ONTOGENESIS 233

relative reactions. This terrestrial type not only differs
from the aquatic type in the absence of gills and in the

5?

o>

00

I

great development of the lungs, but also in the absence
of the crests on back and tail, in the transformation <?f

234 THE MNEME

the rudder tail into a rounded tail, in the shape of the
head and legs, and in the histological nature and pattern
of the skin. When all these distinctive changes have
been finally effected, a return to the neotenic form is
altogether impossible.

Very interesting is the fact, observed by Miss von
Chauvin, that sexual maturity is not attained so long as
the ecphory of the engram-complexes is kept, so to speak,
in the balance between the atavistic and neotenic branches,
when neither the engram-complexes q—q—q, on the
one side, nor the engram-complex Ag on the other side,
become ecphorised. Apparently the engram-component,
the ecphory of which manifests itself by the appearance
of sexual maturity, is associated with the engram-com-
plexes at the ends of both the succession branches, and
can only become ecphorised with either of these. Its
ecphory is thus inhibited, if the course of the succession
is in any way arrested.

It has been mentioned that Miss von Chauvin succeeded
in obtaining, not only at phase ii, but also at later phases,
a passing of the ecphory from the atavistic to the neotenic
branch — say from engram-complex n to k^

This change is first manifested by the arrest of the
atavistic course of development, and then by the re-
gression of the plastic and motor reactions which belong
to the engram-complexes I, m, n, and which have
already occurred. The stump-like gills are restored to
their full size ; the crests, the shrivelling of which had
already begun, stiffen and erect themselves ; the rudder
tail forms itself anew and exercises its function, which
had practically ceased during the atavistic course of
development. After the ecphory of the engram-complex
Aj, the incongruities arising in the homophony between
the now dominating mnemic excitation-complex A, and
the overpowered, but still-existent, original excitation-
complex n slowly disappear.

In Amblystoma a relatively strong external influence
is required to turn the course of ecphory from the neo-
tenic to the atavistic branch. Without such an influence.

ENGRAM DICHOTOMY IN ONTOGENESIS 235

the course of development in the majority of individuals
whose parents belong to the neotenic branch follows
simply in the same branch. But in those Salmandrinae
where a neotenic engram Une has never developed the
course takes place in the usual succession of engrams, and
the engrams of the metamorphosis {I— q in the diagram
on page 233) are generally ecphorised, even when the
external stimuli, which, as original stimuli, on the ancestors
generated the engrams of the metamorphic phases, fall
away. Thus the larvae of Salamandra maculosa accom-
plish their metamorphosis — although late — even if they
are kept in water rich in oxygen, and prevented by a
wire netting from coming into contact with atmospheric air.
The metamorphosed animals would become asphyxiated
if they were not then allowed to emerge from the water
(see von Chauvin, op. cit., 1885, p. 385). The force of
the successive association is here stronger than the preserice
or absence of external stimuU. With the Tritons, however,
an original — ^not a mnemic — neoteny may under special
circumstances be produced, so that these newts are able
to reach sexual maturity while they are still in the gill
stage. By breeding the offspring of such individuals, it
might be possible in the course of time to transform the
original neoteny into a mnemic one ; but as the experi-
ment has not yet been tried, we do not feel justified in
assuming the result. Miss von Chauvin's observation,
however, is a matter of fact (see p. 131), i.e. the offspring
of those Amblystoma, which have grown sexually mature
in the atavistic branch, turn into the atavistic branch
under less external pressure, and pass much more quickly
through the atavistic succession-hne than do the offspring
of neotenic parents. The freshening up of the atavistic
engram Mne in the parents thus effects an easier ecphor-
abihty of this engram-succession in the progeny.

Quite recently a case has been experimentally discovered,
which is remarkable for the fact that the branching off
into the atavistic line, effected by external influences on
the parents, exercises a particularly striking effect on
the progeny in so far as, if not in the first generation,

236 THE MyrEME

then all the more in the later generations, characteristics
recur which had disappeared completely from the genera-
tion on which the experiments were made.

It is well known that most frogs and toads make for
the water when they intend to generate their kind. While
in the water, the male clasps the female, and during many
hours of hard labour squeezes the eggs from her body
by the pressure of the fore-Hmbs. As the eggs, small
in size but great in number, pass out, they are fertilised
by the male. The eggs are covered with a glutinous
layer which immediately swells up in the water and forms
a clear, round, somewhat elastic globe, slippery to the
touch. The eggs, united into masses or into long bead-
like strings, remain Ijdng in the water without any further
attention from the parents. After some time the larvae
emerge, provided with a rudder-like tail. They pass the
tadpole period of their life in water.

A remarkable advance in parental care is shown by a
European species of toad, Alytes obstetricans. With the
Obstetric Toad, the fertilisation and deposition of the
eggs take place on land. The number laid is far smaller
than with other Batrachians, but the eggs are relatively
large and light in colour, and contain more yolk. The
male assists in the process, not only by squeezing the eggs
out of the female, but also by pulling and pressing them
with his hind legs. In the absence of water, the glutinous
envelopes of the eggs do not swell up nor lose any of their
stickiness. On the contrary, they adhere to the hind
legs of the male, and, in consequence of his continual
movement, the egg-strings are gradually wound about
his legs. A few hours later the viscous envelope hardens
and shrinks and loses its stickiness, but the spawn-strings
remain closely adherent to the legs of the male. If the
spawn-strings do not adhere at once, the animal tries by
repeated twisting round to give them the right position.
The eggs safely deposited, the male Aljrtes burrows into
the damp earth until the larvae emerge.

By external influences, Kammerer in a simple manner
induced the Obstetric Toad to return to the more original

ENGRAM DICHOTOMY IN ONTOGENESIS 237

propagation habits of ordinary toads and frogs. He kept
the animals in a room at high temperature (zs'-so" C),
until they were induced by the unaccustomed heat to
cool themselves in the water-trough placed at their dis-
posed. Here the male and female found each other, and
the clutching of the female by the male and the fertilisation
and deposition of eggs took place in the water. Under
these circumstances, the glutinous envelope, coming into
contact with the water, at once swelled up and lost its
stickiness. Thus it was rendered impossible for the
male to affix the spawn-strings to his hind legs. After a
number of fruitless attempts on the part of the male, the
egg-strings were left in the water to develop in the ordinary
way.

By the repetition of these experiments during several
mating periods, the animals gradually accustomed them-
selves to copulate in the water, and to deposit the eggs
without any attempt by the male to twist them round
his legs. Finally, when the coercion of the high tempera-
ture was entirely withdrawn, the creatures went through
the whole process of sexual intercourse in water of normal
temperature. The number of eggs, however, increased ;
the eggs themselves became smaller, poorer in yolk and
darker in colour, until at last the resemblance to the eggs
of the common frogs and toads was almost complete.

Now just as this modification of the propagating and
breeding instincts in the parents grows into a settled
norm, so it appears in their progeny. Kept at normal
temperature, the sexually matured offspring of such
parents make for the water at the approach of their first
breeding period, and deposit their strings of numerous
small and dark eggs without attempting to take any
further care of them. In the ontogenetic development
of this generation, and still more in the case of later
generations, reversions into the ontogeny of the more
primitive tailless amphibians take place ; but into that
we shall not here attempt to enter.

We shall concern ourselves simply with the appearance
of a group of characters which, in the fully developed

238 THE MNEME

Alyies obstetricans, seehied to have entirely disappeared.
The males of frogs and toads copulating in water possess,
as secondary sexual characters, peculiar breeding pads,
which enlarge during each breeding season. These pads
are so characteristic that their location and configuration
are very important for purposes of classification. Their
biological value consists in making it possible for the male
to embrace the female in the water, and to this end a
morphologically very pronounced hypertrophy of the
forearm-musculature appears, in consequence of which
the shape of the limbs assumes a very characteristic,
inward curve. These secondary sexual characters are
absent in the normal male Al3rtes copulating on land.
At any rate, they were entirely absent in the many hundreds
of specimens used by Kammerer in the elaboration of
his experiments. When Kammerer coerced the animals,
in the manner already described, to copulate and to
deposit their eggs in water, no indication of these breeding
pads appeared even when this mode of propagation became
usual. No sign of the pads appeared in the first progeny
bred under those conditions. But in the second progeny,
roughnesses appeared on the thumb and on the thenar
eminence ; and in the third progeny, all the sexually
mature males were furnished with typical, grey-black
coloured pads on the upper side of the thumb and on the
thenar eminence. At the same time there was an ex-
cessive development of the musculature of the forearm,
and with this an inward curvature of the fore-limbs,
which meant that the palmar surface of the hand was
pressed on the ground nearer the median line.

Here we have exceptionally convincing evidence for the
fact that, under the influence of stimulations continued
through four generations, a branch of an engram-succes-
sion long obsolete is brought again into use. Primitive
morphogeneous engrams are reawakened, and their ecphory
is manifested in new reactions. Morphologically, the
case is one of the reappearance of an atavistic character ;
viewed as the manifestation of instinct, it is an
instance of the substitution of the neotenic obstetric

EN GRAM DICHOTOMY IN ONTOGENESIS 239

instinct by the atavistic one of copulation and the de-
position of eggs in water.

In conclusion, the attention of the reader is directed
to the fact that, in all cases quoted in this chapter, the
course of the ontogenetic alternative depends entirely and
without exception on the strength of external influences.
These are open to close analysis, whether the course of
the ontogeny is in the direction of dimorphism or poly-
morphism, of neoteny or atavism. This fact of the
possibility of analysis has been my guiding principle in
the selection of illustrative material.

CHAPTER XIII

THE ORIGIN OF ONTOGENETIC EN GRAM-
SUCCESSIONS

I. The Origination of Dichotomies by Stimulation.

If by stimulation a new link is added to an already
existing succession of engrams c — d — e, the addition
may be by either of two ways. Either the line is con-
tinued so that the hitherto ultimate link e becomes the
penultimate link of the unilinear succession c — d — e—f,

or a dichotomy is affected, c-dC,'va. which the new link

y

/ enters as an alternative to what was previously the
ultimate link e. In Chapter IV I have already analysed
such dichotomies in the sphere of individually acquired
engrams, but without regard to their inheritability.

It is of the very nature of engraphic stimulation that
already existing engrams are never remoulded, but remain
as they were first imprinted. New engrams are deposited
as detached and fresh creations. This characteristic of
engraphic action is the essential element in the formation
of the large group of alternatives in the sphere of individual
acquirements, as also in that of hereditary engraphic action.

As regards individually acquired engrams, the reader
is referred to the experimental and other evidence in
Chapter XV of Th& Mnemic Sensations. In the case
of hereditary engrams, the preservation of the relatively
independent engram is seen most clearly in such phenomena
of atavism as come under our notice in the second part

of the preceding chapter. The old engrams, which had

340

ONTOGENETIC ENGRAM-SUCCESSIONS 241

apparently disappeared without leaving any traces, were
still in being, and only required a special external impulse
to revive them and to reopen the old paths. One of
the most convincing examples of a revived engram-stock
was given in the reappearance of the breeding pads in
the third and fourth generations of the Obstetric Toad,
where, as a consequence of external stimuli, the neoteny
that had become a norm was suppressed in favour of a
return into forsaken ancient tracks.

In other fields, also, reversions or atavisms have been
experimentally effected. Under certain circumstances
they appear quite regularly as a sequel to regenerative
phenomena. Reference may here be made to one
particularly clear case, for the exact demonstration of
which in a large series of different forms we are indebted
to Przibram. On the removal of the third maxilliped
of the short-tailed Crayfish, regeneration-formations arise
which correspond in every respect to the ambulatory
limbs. Only in the case of further " sheddings " does
the gradual transformation into a typical maxilliped
take place. In the normal ontogenesis of the Crab, such
an " ambulatory-limb-like " character does not occur.
But maxillipedes resembling the ambulatory limbs occur
all through life in the longtailed Crayfishes, the ancestral
stock of the short-tailed kind. Analogous atavistic
phenomena occur, as Fritz Miiller ascertained in 1880,
in the regeneration of the claws, and especially of the
fifth leg, of Garneela Atyoida. These, however, were cases
of temporary regenerative stages which were gradually
corrected in later " sheddings."

I wish to emphasize here that, in all the cases hitherto
considered, the reversion was in no way caused by cross-
breeding, but only upon the ecphory of latent atavistic
engrams by external stimuli.

In other reversions, however, the return into old and
disused evolutionary tracks is often a consequence of
crossing. Recent research in hybridisation has succeeded
in showing somewhat more clearly than heretofore certain
interrelations in these cross-bred reversions. Yet the basis

16

242 THE MNEME

of the explanations given is always the actual exist-
ence of " ancestral factors," which by themselves are
incapable of manifestation. In certain cases of crossing,
however, special combinations are re-estabUshed, which
arouse these factors into activity so that they manifest
themselves in plastic and motor reactions.

Attention may also be directed to the fact, proved
long ago, that reversion to characters, which for generations
have been lost or modified, occur also in pure bred forms.
Of great importance in this respect are the careful and
accurate observations on the spontaneous atavistic rever-
sions of oats to wild oats, made in Svalof in the nineties.
More recently H. Nilsson-Ehle established beyond doubt
thai these atavisms arise, not in connection with crossings,
but as an entirely " spontaneous " alteration.

We reserve for our next chapter the investigation of
the question whether these factors arise by stimulation
or not. At this point, however, it will be convenient to
consider the dichotomies which result from crossing.

2. Origination of Alternative Dichotokies by
Cross-breeding.

If two varieties or species distinguished from each
other by a clearly definable character are crossed, as,
for example, a white-flowering plant with a red-fiowering
variety, or a rough-haired with a smooth-haired guinea-pig,
or an obstetric toad of a species mating on land with one
of a species propagating in water, the question arises —
which parental path of development will the progeny
follow ? To use the formula of our diagram : — Will
the progeny follow the path d-e oi d-f at the

dichotomy d^ , resultant from the crossing ?

y

In the exposition that follows we are concerned only
with the progeny issuing from such a crossing. In the
nomenclature of modern science, the first filial generation
-s known as the Fi generation, the second generation

ONTOGENETIC ENGRAM-SUCCESSIONS M3

(grand-children) as the F2 generation, the third as
the F3 generation, and so on.

It is clear that no one can with any certainty predict
which of the two paths, d-e or d-f, will be followed by
the products of the crossing, the Fi generation. Indeed,
in keeping with the phenomena of mixed reactions in the
domain of higher memory (p. 167), the chances are that
the alternative is not decided at all in any hard or
exclusive way, but that mixed reactions occur, from
which an ecphory of the engrams belonging to both
branches may be inferred.

Experience shows that in some cases the one, in other
cases the other possibility is realised. If the alternative
is adopted in a clear and decided way, in the sense that
development follows one path exclusively, then we say,
accepting Mendel's terminology, that the structural or
functional character resultant on the following of this path
"dominates" over the character corresponding to thCj
other path. The latter character is termed " recessive." i

It frequently happens, however, that neither alternative
is exclusively chosen, but a compromise is effected by
the co-operation of the diverging paths of development.
The result is a mixed reaction, and the appearance ofj
an intermediate character in all the members of the Fil
generation. This character, however, need not invariably
be an intermediate one. A product peculiar in many
ways may result from the mixture, as, for example, the
metallic blue of the Andalusian fowl, which regularly
appears when the white variety is crossed with the black.
In some cases the character is, however, in its most literal
sense intermediate, as when a red-flowering variety crossed
with a white gives a pink flowering Fi generation, as in the
case of Lychnis and the miracle flower, Mirabilis jalapa.

There are extreme cases of perfect dominance, where no
trace of the recessive character can be found even on
the most e:jf:acting research, as for example, the dominance
of multicolour over white in the flowers of Lathyrus and
Matthiola, and of hairiness over smoothness in Matthiola.
There ^are cases of the appearance of an exactly inter^

244 ' THE MNEME

mediate character. Between these extremes may be found
all kinds of transition-states.

If we now ask what determines the choice, complete
or partial, we shall have to admit that we really know
very little about the matter. The supposition that the
phylogenetically older path is more or less dominant
over the phylogenetically younger path does not stand
close examination, for the younger path is often taken.
Nor can we say that, in the crossing of species of which
the one possesses a character lacking in the other, this-
character invariably dominates.

In general, the dominance, that is, the absolute or
comparative predominance of one of the hybridisation-
alternatives, is a constant phenomenon occurring in nearly
all the progeny of crossed parents. But irregularities and
exceptions are not unknown. The. subject needs further
elucidation by carefully elaborated experiments. That it
is possible under certain conditions to divert the dominance
into this or that direction by external influences has been
proved by Tschermak's experiments with wheat strains
and by those of Vernon, Doncaster, Herbst and Tennent
in the cross-breeding of various species of Sea-urchins,
and by Tower in that of Colorado Beetles. The researches,
especially those of the last three named, give us some
insight into the nature of the influences at work. Those
of Herbst were the first attempt to analyse closely the
character and action of these influences. We are only at
the beginning of our understanding of them, and we shall
have to leave to further research the task of formulating
the results in any definite way.

From what we have said, it follows that in the generation
which arises as the first product of a crossing — the Fi
generation — a dichotomous combination of the diverging
lines of development has been created. The development
in the members of this generation may then follow one of
three ways. There is either absolute dominance where
the movement is exclusively in one path ; or there is
, imperfect dominance where the movement is chiefly in one
path, but the influence of the other path makes itself

ONTOGENETIC ENGRAM-SUCCES^IONS 245

felt : or there is what may be called intermediate inherit-
ance, where the development progresses with approximately
equal and simultaneous actuation of both engram lines,
so that a mixed reaction results, usually assuming an inter-
mediate character.

As a rule, the separate representatives of the Fi
generation behave exactly alike. This similarity of
behaviour holds good almost without exception in cases
where the F2 generation may be differentiated according
to Mendelian rules.

Let us follow the course of the crossings in the succeeding
generations, dealing first with the second filial generation —
F2. To obtain unimpeachable results and to ensure strict
breeding, it is necessary to exclude all foreign elements
in such a manner as to obtain the F2 generation by the
exclusive use of the products of the first crossing, that
is, the representatives of the Fi generation. This is the
line which Gregor Mendel followed fifty years ago, and
along which he reached his epoch-making discoveries.

Just as one character in the Fi generation dominated
over the other so that it alone appeared in each represent-
ative, or, in the absence of such dominance, an intermediate
path was entered upon so that every representative of
the Fi generation became the bearer of an intermediate
character, thus with the F2 generation certain differences
became pronounced, which on closer scrutiny are merely
the varied manifestations of the identical principles. It
may be well to consider first the case of the complete
or nearly complete dominance of the one character over
the other, taking as our example the classic case given
by Mendel. He crossed two varieties of the edible pea,
Pisum sativum, characterised by a well-marked difference
in the colour of the seed-leaf, one being yellow and the
other green. He found that the yellow dominated over
the green, so that in the Fi generation the offspring
possessed all yellow-coloured cotyledons. In the F2
generation, which in this instance had been obtained
from the Fi generation by self-fertiUsation, " along
with the dominant characters the recessive characters

246 THE MNEME

reappeared in their full peculiarity and in an unmistakable
average ratio. Of every four plants from this genera-
tion, three showed the dominant and one the recessive
character." In regard to the third generation raised
from the crossing, and also bred from self-fertilisation,
Mendel proved that those forms which possessed the
recessive character in the preceding generation did not
vary in the second generation in regard to this character,
which remained constant in succeeding generations.
With those of the first generation which possess the
dominant character, two of the three remaining parts
produce offspring which, just as in the case of the known
hybrid forms, bear the dominant and recessive characters
in the ratio of three to one. One part only remains
constant in the dominant character. If we represent this
by a diagram in which d denotes the dominant character,
and r the recessive, we shall have the following arrange-
ment, which will show the typical splitting of a Mendelian
hybrid form.
Crossing of d with r {d — dominant, r = recessive).

Product of Crossing.
Fx generation d

Fa generation d dd r

/ \

/ ' ' . \

F3 generation d d dd r r

/ / \ \

/ / ' ' > \ \

F4 generation d d d dd r r r

In the case of the crossing of characters where the one
does not dominate over the other, but where the Fi
generation presents an intermediate character, as with
the Andalusian fowl, the modification of the theory is
only apparent. We may take in illustration the miracle
flower, Mirabilis jalapa. By crossing a white-flowering
specimen with a red, we obtain an Fi generation, in
which the offspring bear pink flowers. This generation.

ONTOGENETIC ENGRAM-SUCCESSIONS 247

when selfed or inbred, produces a generation — F2 — in
which one-half bear pink flowers, one-quarter pure white
flowers, and the remainder red. Each of these two latter
groups breeds true, producing when selfed only pure
white or pure red flowers. The pink flowers, when self-
fertilised, produce in the F3 generation the same proportion
of one-quarter true breeding white flowers, one-quarter
true breeding red flowers, and one-half pink flowers.
The last, when inbred, produce flowers in exactly the
same proportion as the pink flowers of the F2 generation.
If we denote the white-flowering specimens by a, the
red-flowering by b, and the intermediate pink by ab, we
obtain the following scheme : —

Crossing of a and b.
Fi generation ab

Fa generation a ab ab b

/ ' ::"-::: \

F3 generation a a ab ab b b

/ / ' rr^-z: " \ \

F4 generation a a a ab ab h b b

It is clear that this scheme is practically identical with
that on the preceding page, as soon as pure dominants
and pure recessives emerge on the self-fertilization of ab,
instead of giving an intermediate character

How are we to explain the different behaviour of the
F2, F3, F4, etc., generations as compared with that
of the Fi generation ? The perspicacity of Gregor
Mendel succeeded in discovering the lines on which the
explanation must proceed. We accept in the main his
views, but later we shall have to show why and how
far we deviate in at least one instance of fundamental
importance from the Mendelian interpretation as generally
advanced.

Take the case we have just considered. On referring
to the diagram, we note that in the Fi generation the
character ab, apparently resulting from the mixture of
a and b, appears in each individual. But in the F2

248 THE MNEME

generation the case is altered, and the " intermediate "
appearing only in one-half of the progeny. Of the
remainder, one-half, that is one-quarter of the entire
generation, shows the character a, and the other the
character h. Throughout the generations a and h breed
true. There is no sign of the character ah, we may
therefore regard it as eliminated. How to account for
this elimination will require more definite consideration
later on.

Now, when does this elimination take place ? Mendel
was aware of the importance of this point. Elimination
takes place at the formation of the germ-cells or gametes.
If we take a representative of the Fi generation and
propagate it by a purely vegetative method, by cuttings,
buds, bulbs, etc., the process may as a rule be continued
indefinitely without the emergence of Mendelian divisions.
Very rarely does the vegetative method allow of the mani-
festation of Mendelian characters.

In most cases of cross-breeding, however, where an
F2 generation results from an Fi generation by the
formation of germ-cells or gametes, a segregation of
characters regularly takes place in 50 per cent, of the
representatives of the F2 generation. The constancy
of this occurrence strongly suggests that the different
behaviour of the F2 generation should be traced to the
germ-cells from which F2 originated. We have, therefore,
to assume that in the germ-cells the factors, which later
on determine the appearance of the Mendelian characters
in the course of the ontogenetic development, are so
arranged that in their evolution one only of the two
retains its efficiency. Its fellow or competitor either is re-
moved altogether or its action is neutralised. In any case,
it is eliminated as a factor.

If we denote the factor that determines the appearartce
of the character a by A, and the one that determines
h by B, we can then say that from one-half of the gametes
of the Fi generation A is eliminated, and from the other
half B. The ratios suggest that the principles which
govern the calculus of probabilities hold good in the case

ONTOGENETIC ENGRAM-SUCCESSIONS 249

of this process of elimination. For if we make the safe
assumption that the competitors whose fate is determined
at the formation of the germ-cells have an equal chance,
then according to the laws of probabiUty A will triumph
over B in one-half of the cases, and B over A in the other
half.

When the germ-cells of an individual of the Fi
generation, of a hermaphrodite plant for example, are
crossed amongst each other, what happens ? As we
saw, of fifty female gametes, twenty-five possess the
factor A, twenty-five the factor B. The same proportion
holds good in regard to the fifty male gametes. Now,
according to the law of probabihties, in one-quarter of
the cases the male gametes that possess the factor A
will unite with female gametes bearing the same factor.
The result will be AA. In a second quarter of the cases
the male gametes possessing the factor B will unite with
females bearing the factor B. The result will be BB. In
the third quarter of the cases male gametes with factor
A will unite with female-bearing factor B. The result
will be AB. And in the remaining quarter, the male
gametes with factor B will unite with females bearing
factor A. The result will be BA. If we now consider
that in the overwhelming majority of cases it is in-
different whether a factor is introduced by the male
or the female gamete, it is easy to conclude that AB is
equivalent to BA. The generation-products in our
examples accordingly arrange themselves in the following
mode : —

25 AA 25 AB 25 BA 25 BB

This corresponds to the distribution of the characters
in the F2 generation, as already noted : —

a ab ba b

Those copulation-products or zygotes which possess only
A as an active factor, if mated amongst each other, produce
individuals with the character a. Zygotes bearing simply
the factor B produce the character b. Hybridisation,

350 THE MNEME

at least in relation to the characters here specified, has
not permanently affected the copulation-products. They
are not hybrids as regards these characters, for they are
not cross-mated, but Uke-mated. They are, according to
modern terminology, homozygotes.

But where gametes bearing the factor A unite with
those bearing the factor B, as in cases denoted by AB
and BA, hybridisation takes place, and zygotes, which
are hybrids in specific regard to these characters, are
defined as heterozy gates. The homozygotes AA, mated
amongst each other, breed true, and the same is the case
with the homozygotes BB. But with the heterozygotes
AB and BA, a fresh segregation takes place in the next
generation in conformity with the laws of probability.
The result is expressed in the formula: —

AA AB BA BB

This recurs in each new generation. Readers are referred
to the diagram on page 247.

If we consider in the same way those cases where crossed
antagonistic characters do not give an intermediate
character to the hybrids (heterozygotes), but where one
character {d) dominates over the other one (r), and if
we denote the corresponding factors or determinants by
D and R, it is clear that, at the mating of the gametes
from the Fi generation, each hundred copulation-
products will arrange themselves in the following ratio : —

25 DD 25 DR 25 RD 25 RR

As character d dominates over character r, the following
formula of character sets out this relation. (Compare
the scheme on page 246.)

This means that, of the seventy-five individuals with the
dominant character d, only one-third are homozygotes
(DD). The remaining two-thirds are heterozygotes
(DR and RD). Thus at bottom the proportion of homo-

ONTOGENETIC ENGRAM-SUCCESSIONS 351

zygotes and heterozygotes is the same as in those cases
where the dominance of one character over the other is
not so apparent. The superficial difference is explained
by the fact that the dominance of one character over the
other prevents us from distinguishing in that generation
the homozygotes DD from the heterozygotes DR and RD.
It is only by inbreeding the apparent dominants that
we are able to distingmsh the pure dominants from the
impure.

The discovery of these fundamental data and their
lucid interpretation we owe to the penetrative genius
of Gregor Mendel. His discoveries have been confirmed
by subsequent investigations in other and wider fields.
It is certain that an explanation of the numerical ratios
in hybridisation on any other basis is impossible.

One point only remains in doubt, a point which does not
affect the numerical ratio of Mendelian characters, but
which is of the greatest importance for any ultimate
conception of the structure of organic substance. How
in the formation of the gametes is the " elimination "
of oiie of the factors effected ? Is it a mechanical separa-
tion of the two factors united in the hybridisation, and
a distribution of these factors to two different gametes,
that is a " segregation " in its most literal sense ? Or
is- the undoubted elimination of one of the factors effected
in some other way ? Is it, perhaps, not expelled, but only
put for a time out of action ? Mendel confidently adopted
the obvious and simple assumption of a segregation of the
factors concerned, and of a distribution of these to two
different germ-cells during the process of cell-division.
Many workers have followed him in this view, and with
the less hesitation in that some of the difficulties, which
at first told against it, have in the course of investigation
been since removed.

Nevertheless, a number of real difficulties still remain.
One objection to the Mendelian assumption lies in the
subsequent necessity of regarding each Mendelian factor
or , determinant as an isolable particle morphologically
independent. This consequence appears to me, at any

252 THE MNEME

rate, too far-reaching to be valid. That the segregation
of the Mendelian factors at the formation of the gametes
is a simpler and easier conception than that of a temporary
neutralisation of one factor by the other is not a sufficient
reason why the theory should be accepted.

Perhaps the " elimination " of one factor by the other
depends simply on the advantage of position which under
usual conditions, according to the laws of probability,
falls in one-half of the cases to the one factor, and in the
other half to the other factor. In certain circumstances,
however, and according to definite rules, this advantage
may be counteracted by external influences in favour
now of the one, now of the other factor. Recent experi-
ments, carefully carried out by Tower from quite new
points of view, confirm this. By varying the external
conditions Tower induced a marked dislocation of the
ratios of segregation as compared with the usual Mendelian
ratio. Tower's experiments are the first systematic
attempt to influence hybridisation by varying the external
conditions, and the results obtained are so remarkable
that they call for further investigation into this hitherto
neglected field, and warn us against hastily making up
our minds on the problem of segregation. For it follows
from the assumption of an actual segregation that the
determining factor in cases of dwarf growth, of steriUty, of
immunity against rust, of colour-blindness, of cataract,
of short fingers, or of specific markings, is a definite iso-
lable particle. Useful as such symbolic ideas may be for
purposes of Mendelian research, the materialisation of
symbol? in the shape of morphologically isolable units
appears to me a very dangerous procedure. Other
reasons besides the simplicity of the assumption and
the ease of ^its application are requisite before we are
justified in accepting segregation, in its Uteral sense, as
a fact.

In passing, however, I should like to point out that
the elimination of one of the competitive factors in the
formation of the gametes— that is, what is usually des-
cribed as the Mendehan mode of inheritance — seems to be

ONTOGENETIC ENGRAM-SUCCESSIONS 253

a process which, by no means appears in all cEises of cross-
breeding. This means that a segregation of the characters
in some part of the F2 generation does not necessarily
take place after each crossing. As a rule, such segregation
does not occur when the two crossed forms differ from each
other so much in their constitution that a systematist
would classify them, not as different varieties, but as
belonging to different species or genera. In crossing
different species, contrary to our experience in the crossing
of varieties, the general thing is that neither dominants
nor uniform intermediate types emerge, but forms, denoted
as Pleiotypes, in which appear numerous gradations due
to the mingling of the parental types. As species-hybrids
are usually sterile, the behaviour of the later generations —
F2, F3, etc. — can only be ascertained in exceptional
cases. Occasionally, however, fertile species-hybrids
have been obtained, which, on inbreeding, showed inter-
mediate characters in the successive generations. Ex-
periments have been made in the crossing of butterflies,
of pheasants, and of rabbits with hares. The two silver
pheasant species, Euplocamus nychthemerus and E.
albocristatus, were crossed and the hybrids inbred. For
five generations the hybrids bred true. Leporides-
hybrids, resulting from hares and rabbits, were bred by
Conrad in the agricultural experimental station at Jena,
up to the sixth generation. Throughout they retained a
number of intermediate characters. True breeding species-
hybrids from the vegetable kingdom are also known.
Research in this direction is, however, only in the initial
stages. The cases so far presented need to be subjected
to further analysis, and to control-experiments continued
through longer lines of generation.

In his latest work Tower reports on cross-breeding
experiments made under differing external conditions
on Leptinotarsa decemlineata, L. oblongata, and L. mttlti'
taeniata, three species of the Colorado Beetle. Left to
themselves on an island in the Balsas river, these three
forms eventually produced a new hybrid race, which
showed an intermediate mixture of all three ancestral

254 THE MNEME

types, but in which the oblongaia-decemlineata characters
predominated over the muUitaeniata characters. This
form bred true without segregations and without reversions.
From time to time, however, spontaneous variations
appeared in the breeds, parallel in every respect to the
mutations of Oenothera lamarckiana in the breeding
experiments of De Vries. Under changed .external
conditions, the same treble crossing resulted in hybrids
which showed the intermediate characters of only two
of the parental forms, either decemlineata-oblongata or
decemlineata-muUitaeniata. These hybrids in the main
main bred true, but occasional mutations occurred.

In conclusion, it may be of use to summarise briefly
the latter section of this chapter.

The crossing of two individuals which differ from each
other in a definite character initiates in the first filial
generation (Fi) dichotomies of the kind already described.
In the ontogeny of this Fi generation, either one path
of development dominates more or less completely
over the other, or both assert themselves in a kind of
compromise which manifests itself in mixed reactions,
that is, in the appearance of an intermediate character
or its equivalent. How the choice of the dichotomous
alternative will fall depends mainly on the specific nature
of the competing paths, and, under certain circumstances,
also on the action of external forces, as recent researches
have shown.

On the formation of the gametes of the Fi generation,
the unstable state of the alternative in the Fi generation
becomes stable. In the large majority of cases this
occurs by the putting out of action of one branch of
the dichotomy. The chances of gain or loss are equal
for both competing branches, so that, according to the
laws of probability, in the one-half of the cases one branch,
in the other half the other branch, is neutralised. Though
it is possible, under certain circumstances, to dislocate
this ratio by external forces, as Tower proved, this ratio
is the rule. This explains the numerical relations of
the Mendelian (segregating) inheritance in its simple and

ONTOGENETIC ENGRAM-SUCCESSIONS 255

in its more complicated forms. Extended research has
but confirmed the findings of Mendel.

In rare cases neither branch of the dichotomy neutralises
the other at the formation of the gametes of the Fi
generation, but there is the fixation of a state in which
both branches exist juxtaposed. This juxtaposition
manifests itself by the constant and exclusive appear-
ance of, intermediate characters in the Fa generation,
and in succeeding generations. These intermediate hybrid
characters breed true. It is possible, therefore, to produce
a pure-breeding hybrid form.

The problem — how in MendeUan inheritance the

neutralisation or elimination of one of the two competitive

factors is effected in the formation of the gametes — we

shall leave in abeyance. Mendel and the greater number

of his followers assume that there is a literal segregation,

that is, a separation and distribution of the two competing

elements over the division-products of the parent cells of

the gametes, probably at the reduction-division. But any

explanation, simple and effective though it may seem to

be, which necessarily leads to the assumption of isolable

particles as representative of the characters in the gametes,

must be regarded with extreme caution. For this apparent

solution of the riddle is simple only as long as we ignore

the greatly complicated, but most delicately attuned and

harmonious, co-operation of those assumed isolable particles.

The problems we have here indicated are in themselves

fundamentally important, but it is not necessary, so far

as the purpose of this book is concerned, to discuss them

exhaustively. Their solution we may safely leave an open

question. Of more direct importance to us is the enquiry

into the origin of the determinants. The next chapter

will be devoted to this problem.

CHAPTER XIV

ENGRAPHIC ORIGIN OF THE DETERMINANTS

In the foregoing pages I have tried to show that the new
potentialities of the germ-cells originate as the products
of stimuli or as the residue of excitations, and that they
correspond in all their properties with the somatic engrams
which have engraved themselves during the life of the
individual on the irritable substance of the body. The
reader is referred to the many examples quoted in the
course of the preceding chapters. To these more might
be added from those submitted in The Problem of the
Inheritance of Acquired Characters. When we consider
that experimental research-work in this direction has
been carried on systematically only during the last few
years, the number of examples available is all the more
remarkable. Experiments by Chauvin, Kammei^er,
Standfuss, Fischer, Schroder, Pictet, Tower, Summer,
Przibram, Blaringhem, Klebs, Bordage, and many others,
have concurrently proved that the new potentialities
originate from excitation.

The question arises whether these engrams are produced
by the direct action of external stimuli on the germ-
cells, or by the conduction of the transformed stimuli
through the irritable substance of the soma. Do . the
engrams result from " parallel induction," or from " somatic
induction " ? The question is discussed in Chapter V of
this book (pp. 129-136), and more exhaustively in a separate
chapter of my essay on The Problem of the Inherit-
ance of Acquired Characters. I think that I have there

demonstrated the insufficiency of physical and physiological

356

ORIGIN OF DETERMINANTS 257

evidence for the general applicability of parallel induction.
But whatever position is taken upj or whatever mode of
induction is assumed, all agree that it is a matter of
induction, and that means, of course, stimulation. No
doubt can exist, therefore, as to the engram-nature of
these new potentiaUties. Tower and Kammerer investi-
gated the action of such newly-generated determinants
or hereditable engrams and found that, if their bearers
were crossed with individuals not possessing them, these
determinants or engrams were inherited alternatingly,
following the Mendelian mode. They behaved exactly
as those inherited engrams with which cross-breeding
experiments are usually conducted.

We have already seen (p. 135) that a stimulus affecting
the irritable substance of an organism and inducing exci-
tation therein, may in some cases affect also the germ-
cells, leaving demonstrable engrams behind, but that in
other cases the germ-cells are unaffected. This apparent
capriciousness of inheritance may be traced to various
causes ; primarily, of course, to the nature and the strength
of the stimulus applied. Recently, however, Tower dis-
covered and fully described a second modifying factor
which in many cases is of decisive importance. He found
in the Colorado Beetle [Leptinotarsa) a temporary period
of extraordinarily heightened susceptibility of the germ-
cells. This we may call the germinal sensitive period.
If stimulation takes place during this period, engrams
capable of manifestation are generated, but the same
stimulation apphed at a time when the susceptibiUty
of the germ-ceUs is normal produces no germinal engraphic
effect. It is possible, of course, that in some cases where
an engraphic alteration of the germ-cells occurs, other
factors may have co-operated to determine the issue.
Our investigation, however, of this problem has hitherto
only gone a short way.

In this connection it may be well to enquire briefly
into the apparently spontaneous appearance of new
potentialities as manifested in the sudden and more
or less isolated appearance of " sports." At the very

17

2S8 THE MNEME

commencement of his researches into the problem of
descent Darwin turned his attention to this phenomenon
of mutation or saltatory variation. Later, De Vries,
working on Oenethera lamarckiana, believed he had dis-
covered the real cause of the origin of new species in
apparently spontaneous mutations. Tower's breeding
experiments have recently furnished important contri-
butions to the subject. The riddle of the apparent
spontaneity — for no thinker would admit the possibility
of real spontaneity — will in a number of cases presumably
find its solution in the varying sensitivity of the germ-
cells. These in certain circumstances become abnormally
susceptible to engraphic action. If by chance an especially
strong stimulus affects the germ-cells just at their period
of heightened susceptibility, the resultant is likely to
be a " sport," a mutation, or a saltatory variation.
Hybridisations appear to create special predispositions
in this direction, as Darwin inferred from the experiences
of breeders. Tower's latest experiments seem to confirm
this idea. In regard to such hybridisations, the possibility
must not be lost sight of that many saltatory variations
may be simply the result of new groupings of already
existing engrams. They do not represent an3rthing
really new, especially when the variation is of a regressive
or atavistic nature.

Attention must here be called to a noteworthy peculi-
arity of many newly formed " mutations." As shown in

the preceding chapter, an alternative character dC may

7

be generated by stimulation (p, 240) as well as by
crossing (p. 242). In this latter case, dichotomy appears
in the heterozygotes of each generation. On crossing
these again with either of the two parents of the
cross, the heterozygotes distribute themselves in the
ratio I : I in conformity with the Mendelian law. The
same ratio, however, is also revealed by a good many
true breeding mutations, if these are crossed with the
original form from which they sprang. They behave

ORIGIN OF DETERMINANTS 259

like heterozygotes. Thus in them an alternative exists
which, latent in pure breeding, is Manifested on recross-
ing. In mutations which do not breed true — a case by no
means rare — ^the alternative already manifests itself in
in-breeding.

When we speak of the genesis of the determinants
by stimulation, that is, by engraphy, we do not thereby
wish to convey the idea that the single " determinant "
or " gen " always corresponds exactly to one engram.
Recent research in heredity has proved in a great number
of cases that the determinant, which one was at first
inclined to regard as simple, is in reality exceedingly
complex. Such complex factors, therefore, are not to
be regarded as single engrams, but either as engram-
complexes or as co-operating engrams of different origin.

Again, what appears to be a single stimulus is usually
a very complex phenomenon. The seemingly simple
sunbeam, for instance, that strikes our face is composed,
according to the wave-length, of chemical, visual, and
thermal rays which, acting on different stimulus receptors,
simultaneously originate different kinds of excitations
and leave behind different kinds of engrams. The same
ray stimulates in various ways the plant on which it
impinges, and generates engrams of various kinds. Sumner
and Przibram reared young mice and rats in an unusually
high temperature, and found that the external skin was
the first thing affected. The regulated bodily heat of
the animals modified in a great measure the influence
of the atmospheric temperature on the rest of the organs
of the body. The skin and its glands and, above all,
the hair were changed under the abnormal conditions.
Corresponding hereditable engrams were left behind
in the germ-cells, as was shown by the fact that the off-
spring of these animals, even when reared in a normal
temperature, exhibited the induced alterations. Abnorm-
ally high temperatures act also on the entire metabohsm
of the animals, accelerating development, diminishing
growth, and producing premature sexual maturity. The
excitations producing these reactions leave behind in

26o THiE MNEME

the gerni-cells hereditable engrams which determine the
corresponding alterations in the offspring. Thus, the
same thermal stimulus, acting on the complex organism,
causes different kinds of excitations and leaves behind
different kinds of engrams. These latter, it is true, are
simultaneously associated, but there is the possibility
that by properly arranged crossings the association or
correlation may be broken up. Cross-breeding has
proved itself in many cases to be instrumental in break-
ing up correlations. Whether in the case just discussed
this holds good will have to be ascertained by further
investigation.

There are also, however, factors of quite a different
origin, engrams which were produced at distinct times
in the individual or in the ancestral line, and which,
therefore, cannot be defined as engram-complexes according
to our meaning. These can determine characters which
seem to be throughout of a uniform nature, but regarding
which evidence can be adduced-^ — especially from the
analysis of crossings — ^that they arise by the co-operation
of different determining factors. It may be well to con-
sider in detail some of the results of recent experimental
research, which has disclosed certain characteristics of
the hereditary determining factors ; these characteristics
correspond in a most striking way with those already seen
to belong to individually acquired engrams.

It has previously been pointed out (p. 169) that if a
stimulus which has already generated an engram again
acts on the individual, it does not reinforce or intensify
the existing engram, but creates a new engram similar
but distinct, and ecphorable separately. If two or more
equivalent but isolated engramS are ecphorised, then
homophony of the corresponding mnemic excitations
follows. If these homophonous excitations manifest
themselves by reactions of feeHng, the excitation result-
ing from the homophony is strengthened not so much
in its intensity as in its vividness.

What interests us here especially is the fact that the
repetition of the same stimulus generates in an individual

ORIGIN OP DETERMINANTS 261

a number of qualitatively equivalent engrams, which main-
tain their separate identity. They are distinct entities,
and yet they display a joint homophonous activity.

Nilsson-Ehle has recently demonstrated in his interesting
crossing experiments with oats and wheat, that in many
cases a character owes its development to the co-operation
of a number of qualitatively equal determinants or " units."
Thus, his investigations (p. 66) gave " for the brown
colour of the ears of wheat, as well as for the black colour
of the beard of oats, several units, or at least more than
one, which could not be distinguished qualitatively from
each other." He further showed (p. 71) that " the red
COLOUR OF Swedish velvet wheat is determined by

THREE UNITS, INDEPENDENT OF EACH OTHER AND EACH
DIVIDING BY ITSELF, AND THAT EACH OF THESE UNITS
IS CAPABLE BY ITSELF OF PRODUCING THE RED COLOUR."
. . . "It can BE STATED WITH CERTAINTY THAT THE
DIFFERENCES IN THE MANIFESTATION OF THE DIFFERENT
UNITS FOR THE RED COLOUR OF THE CORN ARE HARDLY
PERCEPTIBLE, AND THAT THERE CAN CERTAINLY BE NO
QUESTION WHATEVER OF A QUALITATIVE DIFFERENCE

IN THE UNITS, either in those for the black colour of the
beard of oats or in those for the colour of the ears of
wheat." The same comment is valid for other character-
istics, such as the panicle inflorescence, the beard in
wheat, and the lingule-character in oats. " No difference

IN THE CHARACTER OF THE LINGULE IS DISCERNIBLE AS
RESULTING FROM THE TWO INDEPENDENT LINGULE-UNITS
OF THE KIND O353, AND THEREFORE THE EXTERNAL APPEAR-
ANCE GIVES NO CRITERION OF THE PRESENCE OF ONE OR
OF THE OTHER UNIT" (p. 89).

Nilsson-Ehle showed in his work the important bearing
of this peculiar fact on the theory of descent, and how
the appearance of certain saltatory variations, as also
the manifestation of inherited gradations of certain
qualities, might be deduced from it. It is not impossible
that many cases of seemingly constant intermediate
inheritance might also find their explanation on the
same basis of fact. Sufficient evidence in this direction

262 THE MNEME

has not been adduced yet, but, so far as the subject of
this book is concerned, the important point is that we
have now indubitable evidence of the co-existence in
the same individual of hke determinants, that is, in their
manifestation.
Nilsson-Ehle by his cross-breeding experiments

HAS THUS arrived AT EXACTLY THE SAME CONCLUSIONS IN
respect to THE HEREDITARY FACTORS AS I HAVE REACHED
BY AN ALTOGETHER DIFFERENT WAY IN RESPECT TO
INDIVIDUALLY ACQUIRED ENGRAMS AND THE CO-EXISTENCE
OF ISOLATED BUT QUALITATIVELY SIMILAR POTENTIALITIES,
EACH OF WHICH ON ECPHORY CAN OF ITSELF PRODUCE THE
CORRESPONDING REACTION, BUT THE HOMOPHONOUS CO-
OPERATION OF WHICH BRINGS ABOUT UNDER CERTAIN
CONDITIONS AN INCREASED EFFECT.

On this, as well as on other points, the method of reason-
ing adopted in this book and the conclusions reached in no
way conflict with the result of researches on variation
and hybridisation ; rather are they supplementary. The
advances which we owe to these modern branches of
scientific research are incalculable. A mathematical
precision, hardly to be obtained elsewhere in the domain
of biology, distinguishes them and makes their study
an sesthetical pleasure. The fact, however, must not
be lost sight of that this line of research regards the
determining factors either as definite magnitudes, or
as symbols detached from the characters. The symbols
are used not only as such, but also as if they possessed
the marvellous capacity to behave at all times and in all
places in the manner prescribed.

In proceeding thus in regard to variation and hybridisa-
tion research has followed the only right method. It has
thereby very nearly reached a solution of the problem
presented.

The limitations of this method, however, which is
fully justified in its own proper field, must be borne in
mind, otherwise our conclusions might overstep the right
limits of this kind of research. We must not ignore, for
example, the physiological aspect — ^how the " determin-

ORIGIN OF DETERMINANTS 263

ing factors" fulfil their proper task and function—
or leave out of account the data accumulated by ex-
perimental embryology, such as the phenomena of
regeneration and of periodicity, that is, the co-operation
of the determinants.

To my mind the facts do not warrant the inference
that the determinants are isolable particles of matter ;
such inference would ohly be valid on the supposition
that in no other way can " segregation " be explained.
The question, however, is in our present state of knowledge
too difficult to be handled here. Although personally I
take an entirely different view from that just mentioned,
I prefer to leave the matter open.

The isolability of the structure, known as the determin-
ing factor, is a subject worth discussing. One conclusion,
however, accepted by many biologists is quite unwarrant-
able : that a lack of correlation of these structures amongst
themselves has been proved, and that we have therefore
a right to assume that the determining factors were
Isdng about in the germ-plasm in a disconnected and
unorderly way.

The biologists referred to were led to this position by the
fact that they were able to spUt up character-correlations
established by cross-breeding, and to separate characters
which were apparently indissolubly united.

Against this we may set the fact that correlations
exist, the segregation of which by crossing has so far
not been accomplished. The reader is reminded of the
strange phenomenon of gametic union. Further, cases
exist where two factors cannot be made to unite by
crossing. They seem to repel each other. It appears,
however, not improbable that such cases of spurious
allelomorphism may yet be explained. Nilsson-Ehle,
for instance, showed quite recently that, under certain
circumstances, one of the factors may act as a check,
and so neutralise the action of several quite independent
factors. In that way factors are correlatively united,
and the relation may be determined by factors acting
independently and in no sort of special unity. But

264 THE MNEME

even if we assume that it is possible by cross-breeding
to separate the various factors from each other, this
does not at all prove that in the antecedent state there
was a complete lack of correlation among these elements.
Latterly the analyses of those engaged in hybridisation
research have been frequently compared with chemical
analyses, a comparison against which many objections
might be raised. Allowing, however, the comparison,
we may note that, in any chemical combination, an
orderly arrangement of the respective units and a definite
union under the image of more or less closely coherent,
subordinated complexes is assumed. We may think, for
instance, of the six carbon atoms in the Benzene ring,
and of the subsidiary attachments of the six hydrogen
atoms. Nobody dreams of asserting that the possibihty
of separation implies a non-orderly juxtaposition of the
chemical units.

The segregations brought about by hybridisation are,
in my opinion, essentially different from the separations
of chemical analysis. But granted that I may be entirely
mistaken in this view, it by no means follows that a
differently graduated relation, an established contiguity of
the genetic determinants, such as is generally assumed
in this very chemical parallel, does not exist. The exist-
ence of such contiguity seems to me to be proved by
the physiological data of development, of regeneration, of
regulation, and of periodicity.

We also find individually acquired engrams in a well-
established contiguity, which on ecphory manifests itself
in the phenomena of simultaneous and successive associa-
tion. The simultaneous engram-complexes form, it is
true, a coherent unity, but we can nevertheless produce
from the single components of these complexes entirely
new combinations. In Chapter IX of The Mnemic
Sensations I have defined these processes as " the associa-
tion of the components of different engram layers." The
manner in which, in the individual life, the components
pf engram-complexes are used over again for the creation
of new complexes is markedly different from the process

ORIGIN OF DETERMINANTS 265

involved in the re-combination of the components during
hybridisation. But the " tertium comparationis " is
contained in the fact that in both cases a severance of
the original associations and a new combination of the
components are possible, and that it would be illicit
in both cases to infer from this fact that no established
contiguity of the components had ever existed.

The connection of the hereditarily transmitted " factors,"
their co-operation, the localisation of their action in time
and place, are problems of such a nature as to be incapable
of solution by the methods of MendeUan research alone.

We may quote Tower's just observation, " To say
that together they direct the development of one part after
another in orderly succession puts upon these determiners
a biirden of great responsibility — almost involving in-
teUigence — and makes necessary some co-ordinating
mechanism behind it all."

If we acknowledge the engraphic origin of the determin-
ants — a position proved by experiment in many cases —
and if we use the two principal mnemic laws as keys to
all the processes determined by the ecphory of engrams,
then by grouping the accumulated data on hereditary
and on individually acquired engrams under common
heads, we obtain some idea of the " co-ordinating mechan-
ism " of which Tower speaks, and we do this without
leaving strictly physiological ground and without having
to ascribe anything like intelligence to the determinants.
Our procedure of deducing the orderly arrangement in
time and space of ontogenetic processes from a law of
a more general order thus furnishes the necessary supple-
ment to the methods of Mendehan research, which has
primarily altogether different aims. A bridge is thereby
thrown across from the study of variation and hybridisa-
tion to that of experimental embryology, and. it is made
clear that the action of the determinants is im accord
with purely physiological laws. The methods of this
procedure are markedly mechanistic, and its results
render the assistance of any vitahstic principle entirely
superfluous.

CHAPTER XV

THE PROPORTIONAL VARIABLENESS OF
MNEMIC EXCITATIONS

In the chapter on mnemic homophony we explained
that the mnemic state of excitement is a repetition of the
original state in all its proportional values. But let
us here emphasise that it is so in its proportional values
only, not in its absolute values. According to the energetic
condition at the time of the repetition, the mnemic
excitation or succession of excitations may be less or more
vivid than the creative, original excitation. Moveover,
from the same cause or from the influence of new original
stimuli, the mnemic repetition may take place in a tempo
slower or quicker than that of the original excitation, but
the ratio of the time periods, that is, the original rhythm,
will be uniformly maintained.

If a change in the internal or external energetic condition
at the time of the reproduction causes a proportional
change of the mnemic excitations, this change may affect
all values which can manifest themselves by reactions
differing in quantity.

We have the power, as we well know, to give what
size we like to a mental image when projecting it spatially.
The engram of any form can be ecphorised either greatly
magnified or greatly reduced, according to the nature
of ecphoric influences, of homophonous original excita-
tions, or of co-operating associations. An example of this
may be seen in the case of a sculptor or painter who is
able to reproduce an original percept in different dimen-
sions, but with perfect fidelity to the proportions. Most

366

PROPORTIONAL VARIABLENESS 267

people unconsciously write smaller between narrow lines
than between wide lines, yet, proportionately, the
letters are a correctly reduced copy of the usual and
characteristic hand of the writer. Proportional variation
in size of handwriting may also occur, mainly as motor
reaction, when writing with the eyes shut.

In like manner a succession of mnemic excitations
may proceed more quickly or slowly than the original
excitations, without disturbing the original proportions
between the members of the succession. Under the
influence of a conductor beating time, or 'of a fellow
singer, or of a pianoforte accompaniment, or of emotions
heightened by the stimulus of alcohol, a piece of music
may be sung, consciously or unconsciously, in a much
livelier tempo than previously.

These two examples, the changed size of handwriting
and the altered tempo of a piece of music, are characterised
by the fact that, while the proportional alterations of
the mnemic processes may be accompanied by reactions
in consciousness, this does not necessarily follow. Nor
is it solely the peculiarity of engrams acquired during
individual life that their mnemic excitation is affected
by the energetic condition ruling at the time of ecphory,
the alteration still maintaining the original relative
proportions, for we meet with a like peculiarity in the
case of inherited engrams, and this in a very striking
manner.

Just as the tempo of a piece of music may be changed
by any of the above mentioned influences, so the tempo
of the ontogenetic courses may be retarded or accelerated
by lowering or raising the temperature without any
alteration of the proportions of the rhythm, which is often
extraordinarily complicated. For example, the tempo
of the morphogenetic processes in the egg of the frog
is more than four times quicker if they are initiated at
a temperature of 24° C instead of 10" C. But in both
instances the rhythm of the processes is identical. Again,
if in a developing organism the quantity of material at its
disposal for plastic processes is reduced, or if the production

268 THE MNEME

of fresh material is hindered, the entire plastic structure
is built up in proportionately reduced dimensions, and
the resulting organism is relatively smaller than the normal
type. Conversely, if the quantity of the building material
is increased or the abnormal production of fresh material
is induced, everything is developed in proportionate en-
larged dimensions. The organism becomes enlarged as a
whole, but in proportion strictly corresponding to the
normal type. It is experimentally easy to obtain a
decrease in building material by severing the egg of the
animal during the process of segmentation. Medusae,
Echinodermata, Acraina, - etc., provide fit material for
observation. Or, during their development, we may de-
prive the plants and animals of adequate nourishment,
or otherwise place them in unfavourable conditions of
life, and the result will be a decrease of material for the
building up of the body.

On the other hand, an increase in material may bq
obtained by taking two ova at the blastula stage of
segmentation, and inducing them to blend into one
organism as Metschnikoff succeeded in doing with Medusae,
and Driesch with Echinodermata. Or, by cultivating
organisms under very favourable conditions of life and
by nourishing them richly, we may succeed, especially
with plants, in obtaining a proportional enlargement of
the whole organism, or of single parts such as flowers or
leaves, far beyond the normal dimensions.

Cases might be multiplied without difficulty and extended
to the various regions of mnemic operations. They prove
that this phenomenon of proportional rhythms is specifi-
cally characteristic of all mnemic excitations, and of all
successions of such excitations.

In organisms of restricted growth, where the proportions
of the parts are definitely determined engraphically, the
disturbance of the proportions produces an incongruit3'
of the mnemic homophony, and induces reactions which
tend to remove the incongruity. Animal life furnishes the
greater number of instances. Let us take, for example,
the well-known case studied in detail by Th. H. Morgan.

PROPORTIONAL VARIABLENESS 269

If, from a Planaria, a segment is cut so small that it
contains only the material for the building upi of an
animal one-fifth the size of the original worm, and if,
in the segment an organ, say the pharynx, is left intact,
this does not retain its normal size, but as soon as the
processes of regeneration begin, it is, so to speak, melted
down and rebuilt in proportions proper to the organism
as a whole. It is evident that, in the proportionately
reduced whole, the original excitations due to the presence
of the unreduced pharynx introduce a strong incongruity
into the homophony of the mnemic excitations, which
correspond to the reduced whole. This incongruity
is removed by the reactions involved in the absorption
of the old, and the reconstruction of a new pharynx.
This remarkable phenomenon, one of the most marvellous
instances of regulation we know, ranges itself, on the
theory of the proportional variableness of mnemic excita-
tions, amongst that large group of reactions which tend
to remove incongruities in the mnemic homophony.

We may now state comprehensively that the engram
determines, not the absolute quantity of the mnemic
excitation resulting from its ecphory, but only its quality
and its ratio to other mnemic excitations associated
with it.

In The Mnemic Sensations 1 have investigated in detail
what determines the absolute values of the mnemic excita-
tions aroused at each step in the ecphory of a succession
of engrams, at least, as far as the excitations manifest
themselves by reaction in sensation. I have reached
the conclusion that, in consequence of the engraphic
fixing of certain additional characters, the reproduction
normally occurs as to space, time, and intensity, in the
same values as in those of the original sensations, but
the proportional increase or diminution of these values
has to overcome a certain, however sUght, inertia. The
reader is referred to Chapter XIII of the above-mentioned
book for further particulars.

In mnemic excitations which manifest themselves
by plastic reactions the same principles hold good. In

270 THE MNEME

this area, also, the determination of the absolute values
depends, in the first place, on the co-operation of certain
additional engraphically fixed factors. The number of
cell divisions, which have to be made before a definite
new developmental phase can enter (see p. 75), appears
to be the first thing engraphically determined. Thereby
a certain norm for the absolute size of the organs at the
various stages of development is given. It is true we
have already admitted (p. 77) that this factor is not
the only determining one, but that under certain circum-
stances other factors may play a still more decisive part.
This only proves that, in this area too, the resistance
against the enlargement or decrease of the normal typp
can be overcome by various internal and external factors,
although only within certain mnemically fixed limits.

The absolute values for the reactions resulting from
the mnemic excitation are thus determined in the first
place by additional components — engraphically fixed —
of hereditary or of individual origin. These become
simultaneously ecphorised, and determine according to
their nature certain absolute values. In the second
place, the absolute values are determined also by original
influences, active at the time of the ecphory. As we
observed, on the appearance of well-marked incongruities
the homophony acts in this instance also as the regulating
factor.

PART IV

CHAPTER XVI

RETROSPECT— REJOINDER TO CRITICISMS

So far my purpose has been to furnish evidence of a
common physiological foundation for the apparently
heterogeneous organic phenomena of reproduction. The
two corner-stones of this foundation are, first, the fact
of the primary coherence of all simultaneous excitations
in the organism ; and second, the fact that the effect of
the stimulus does not entirely vanish with the synchronous
nor yet with the acoluthic phase, but that after the dying
away of the latter, the " engram " — an enduring material
change of the irritable substance — remains behind, which,
though latent, can be roused to manifestation at any
moment in conformity with known laws. As the germ-
cells are not separated from the rest of the body by any
isolating contrivance, the excitations produced in the
irritable substance of the body reach them also, leaving
behind, especially during the period when the germ-cells
are most sensitive, engrams which later may become
manifest.

The above foundation is formally expressed by two
correlated mnemic laws — the law of Engraphy and the
law of Ecphory.

The First Mnemic Law. or Law of Engraphy.— AW.
simultaneous excitations within an organism form a
coherent simultaneous excitation-complex which acts
engraphically ; that is, it leaves behind a connected
engram-complex constituting a coherent unity.

The Second Mnemic Law, or Law of Ecphory. — ^The
partial recurrence of the energetic condition, which

i8 . '»

274 THE MNEME

has previously acted engraphically, acts ecphorically on
the whole simultaneous dngram-complex ; or, as it may
be more explicitly stated : the partial recurrence of the
excitation-complex, which left behind the engram-complex,
acts ecphorically on this simultaneous engram-complex,
whether the recurrence be in the form of an original or
of a mnemic excitation.

Association is the nexus of the single components of
an engram-complex. The engram-association is a result
of engraphy and becomes manifest on ecphory.

A further result of our investigation is the insight
gained into the nature of Homophony, or the co-operation
of simultaneous and qualitatively similar excitations.
When there is practically complete similarity between
these simultaneous excitations, there ensues a perfect homo-
phony ; when not so, the homophony is a differentiated
one.

In the light of these ideas, I have examined in the
third part of this book the action of the mnemic processes
in ontogenesis. I hope it has been sufficiently demon-
strated that the mnemic laws, without the help of any
other theory, do adequately account for the strictly
regional and temporal order of the ontogenetic processes,
for the entry of each of the innumerable separate occur-
rences in given places and at given times, for the dependence
of all simultaneous processes on each other, and for their
occasional independence. Also with equal cogency we
can bring the phenomena of hereditary and non-hereditary
periodicity within these laws, and further we have the
key to a physiological interpretation of the otherwise
mysterious processes of regeneration and regulation.

Before I examine the objections which have been raised
against the mnemic theory so set out, let me discuss a
criticism which is concerned not so much with its truth
as with its explanatory value and its fruitfulness for
science. Granted the truth of the exposition, " What
else is it but another paraphrase of old enigmas ? " I
am asked.

It will be well to state the possible objections at further

RETROSPECT 275

length, and to reply to them seriatim ; they have often
occurred to my own mind.

First, the utility of the phrase " engraphic action of
stimuli " and of the word " engram " may be called in
question, and it may be asserted that the theory does
not advance matters at all, in the absence of any explana-
tion of the real nature of the engram and of the engraphic
action of stimuli.

I reply that neither science nor philosophy has yet
explained the " real nature " of any phenomenon what-
ever. I claim, however, to have succeeded in discerning
some phases of the engraphic action of stimuU in their
orderly recurrence, and in reducing them to a minimum
of fundamental law, and this too without the aid of any
other hypothesis, and based entirely on actual observation,
which can be tested at any moment, the only exception
to this being engrams not capable of repetition. It may
be, I admit, that the nature of the engraphic action of
stimuli is not explained, but its explanation has in the
" scientific sense " been initiated.

In the first German edition of this book I stated that
I had reUnquished the attempt " to reduce the engraphic
change to a hypothetical dislocation of hypothetical mole-
cules."

This point of view, which was simply the attitude of
cautious reserve, has been altogether misunderstood by
some of my readers and critics, one of whom has reproached
me for giving a metaphysical aspect to the engram-doctrine.
No reproach is less justifiable than this. Throughout I
have carefully avoided the notion that the engram might
be something immaterial or metaphysical. On the con-
trary, I have conceived and definitely described it as a
material alteration. I made the point clear in describing
the engram as a change left behind in the irritable sub-
stance after the excitation has died down. As the altered
state of a substance, the engram must necessarily be sub-
stantial or material, and it may therefore be quite correctly
described as a material alteration.

Moreover, in this book, as well as in The Mnetnic

276 THE MNEME

Sensations, special attention has been devoted to the
specific structural side of engrams, and to their localisation.
In The Mnemic Sensations (pp. 282 and 373) I have
called attention to the existence of a chronogeneous
localisation, an important new factor which must be
borne in mind in the further analysis of the problems of
locahsation.

Later, it will be necessary to refer to the question how
far it is admissible to associate the problems which have
occupied us here with invisible structures whose existence
is a matter of pure imagination. Gra,nted that the mor-
phology of engraphic action must remain unknown until
the more minute structure of the engram is better under-
stood, is it not of great advantage in dealing with biological
problems to replace as far as possible such unknown
quantities as memory in the restricted sense, heredity.
Capacity for regeneration and periodicity, etc., by the
function of one single factor — " the mnemic excitation,"
whose existence becomes the more firmly established
the more one studies its multifarious manifestations ?
If by close study we find that all the apparently hetero-
geneous phenomena may be referred to three fundamental
principles, namely, the two mnemic laws and the laws
of homophony, which in their turn are simply sequels
of synchronous stimulation, have we not made a step
forward in actual knowledge by this simplification of our
conceptions ?

But critics may object that when we describe the
phenomena of regeneration and regulation as " reactions
for the removal of the incongruity of a homophony,"
we have not explained the problem, but merely re-stated
it in other words, since it is just the way in which the
removal of the incongruity is effected which constitutes
the essential problem to be solved, and by sapng that
a reaction appears which removes the incongruity we
are guilty of reasoning in a circle.

To this I reply that I am fully conscious of not having
made sufficiently clear the processes of regulation in all
their essential relations; But, it may be added, their

RETROSPECT %J7

explanation has been raised to an entirely different
level by the introduction of a new factor, namely, the
conception of Homophony.

This conception helps us to realize the existence of
two excitations in the regulating organism — an original
excitation produced by the stimulus which thus causes
a temporarily abnormal state, and a mnemic excitation
which belongs to the normal state of the organism or
its ancestors. How the regulating reactions arise and
run their course under the joint influence of these two
excitations is at present very obscure, and probably will
be until the relations between excitation and reaction
in general have been far more closely studied and more
accurately discerned than at the present stage of our
physiological knowledge is possible.

The problem has become amenable to scientific treat-
ment only since the acquisition of this evidence of the
action and counter-action of two actually existing excita-
tions in the processes of regulation. It was merely a
metaphysical problem so long as one had to speak of
regulation as bearing on some imagined future state.
The discrediting of conceptions which fail to touch the
objects themselves, and their replacement by the idea
of the homophony of two material processes, which
change from phase to phase, but are always there, suggest
at least that the introduction of the mnemic principle
is by no means a repetition in a new form, but a con-
ception which renders the problem soluble by physiological
methods.

The argument for the autonomy of form-development
and of life and the case for Vitalism are founded mainly
on the assumed impossibility of attacking the phenomena
of heredity, adaptation, and regeneration by the way of
mechanical causality alone, and without the assumption
of a special teleologieal principle. Our mode of regarding
these problems renders unnecessary vitalistic principles
like Entelechy, and gives us a vantage ground from which
these problems may be solved in a purely physiological
way.

378 THE MNEME

Let us now turn to those attacks which are directed
against the correctness of our conception, and especially
against the citadel of our position, viz. that all the hetero-
geneous phenomena of reproduction rest upon a common
physiological basis. Of two main objections, the one is
directed against the assertion that characters acquired
during the individual life are not transmissible, the other
against the identity of the laws governing the higher
memory and those governing inheritance, though a
certain similarity between these laws and the possibility of
transmission of acquired characters are both admitted.

The objection that individually acquired engrams are
not transmitted to offspring has been met by established
facts, some of which are given in the preceding pages.
For fuller information and for a critical appreciation of
the problem in all its bearings, the reader is referred to
my essay. The Problem of the Inheritance of Acquired
Characters. Modern experimental research has fur-
nished what to me is incontrovertible evidence that the
new inheritable acquirements of the organism originate
as products of stimulation or induction, in whatever
manner the latter may have arisen, and that they have
to be regarded as engrams. This conclusion is not to be
shaken by the mere assertion that the engraphic altera-
tion of the germ-cells is effected, not by a conduction
of the excitations from the rest of the body to the germ-
cells (somatic induction), but by external stimuli pene-
trating directly to the germ-cells, in a manner equivalent
to their action on the rest of the body (parallel-induction).

For, as my critics must admit that the engrams of
the germ-cells possess exactly the same properties as
those of the soma, which are generated by parallel-
induction, it makes little difference, so far as the engram
theory is concerned, which kind of induction is preferred^
But as regards the effect of external stimuli, I think I
have made it sufficiently clear in the above-mentioned
essay that a great number of physical and a still greater
number of physiological facts render impossible the
general application of the idea of parallel-induction ;

RETROSPECT 379

and, therefore, the engram-theory must proceed on the
assumption of the somatic induction of the germ-cell.
It is on these Unes that the theory in the present book
has been elaborated.

Those who refuse to accept this line of argument, and
who prefer to adhere to the assumption of parallel-in-
duction, an assumption quite inapplicable in many cases,
and from the physiological side easily assailable, might
be reminded that they also assume a stimulation of the
germ-cell, that is, an hereditary engraphy, and that
they, therefore, are obliged to build on the foundation of
an engram theory. Their objections against the Mneme
theory cannot, therefore, be regarded as fundamental.

Let us turn to the second objection, which says that,
even granted the inheritance of acquired engrams, the
case is one of mere analogy with the concurrence of the
phenomena of higher memory and heredity, and that it
is not the manifestation of a common conformity to the
same fundamental laws of what is, in principle, an identity.
That the cerebral process of memory is not identical
with the process of growth in embryonic development,
and that the latter is not identical with the unconscious
movement of the plant — well, " There needs no ghost . . .
come from the grave to tell us this."

For the " memory process " is not identical in the
strict sense with any other process. I may remember
one minute, and five minutes later recall the same visual
impression, but the processes are not identical. Every
organic process, every phenomenon can of course be
identical only with itself, and yet one is justified in
stating that an identical principle may be at the root
of phenomena non-identical and very different to the
outward eye.

A simple water-wheel attached to a lawn-sprinkler
or garden-fountain is very much unlike a turbine of
modern construction, and the processes of the two are
by no means identical when compared in detail ; yet
the physicist will tell us that the principle underlying
both processes is identical, and not merely analogous.

28o THE MNEME

Oscar Hertwig has been quoted as an adherent of the
" analogical " theory. He repeatedly emphasised that
" there exists a ' remote analogy ' between the marvellous
properties of hereditary matter and the no less marvellous
properties of brain matter. That this analogy is no
identity will be patent to the clear thinker." We have,
therefore, often cited O. Hertwig as an opponent of the
theory here developed. It is with peculiar satisfaction I
announce that he can no longer be regarded as such.
For, in the later editions of his book, he expressly recognises
the identity which we affirm. " In my opinion, the
phenomena of the ' Mneme,' that is, of heredity, and
the phenomena of memory fall under the general con-
ception of reproduction, and, thereby, show a certain
identity in their character, a fact I have never disputed
nor do I question now." In the light of this statement,
O. Hertwig, in spite of certain qualifications which he
makes and to which we shall refer later, cannot be claimed
as belonging to those who deny the identity of principle
underlying the phenomena of heredity and of the higher
memory.

As regards the opponents of the possible identity of
the physiological foundations of all organic reproductions,
it would seem that their main difficulty is the diversity
in the form of expression of two or more excitations, a
diversity which forbids the idea of conformity to any
single principle. Nobody will deny that motor reactions
are altogether different from those of secretion, and
both again differ from plastic reactions. The nature of
the consequent reaction, whether motor, secretive, or
plastic, is of no more importance as an indication of the
character of a mnemic excitation than it is in the case
of an original excitation.

Plate, indeed, maintains " that the Mneme is a psychical
process, and, consequently, like all psychical phenomena,
absolutely enigmatic. It is, therefore, unsuited for an
explanation of heredity, that is, for any real understanding
of the physiological processes involved in heredity." It
is interesting to compare this statement with one made

, RETROSPECT 281

by the same author a year earlier, when, quite correctly
according to our views, he wrote : " I do not see what
difference it makes that the case is one of psychical process
in the one realm and of material process in the other ;
for, as the former is indissolubly connected with the
substance of the nervous system, both are radically affected
by changes of the protoplasm, and these can be directly
compared with each other." If this is correct, then it
is wrong to characterise the mnemic processes and their
conformity to fixed laws as psychical. Throughout my
argument I have considered the material aspect, a fact
Plate adequately recognised in 1908. Further, I have
attempted to estabUsh a physiological theory of excitation,
and have used the psychical side of phenomena as mani-
fested in sensations, as one out of many characteristic
ways in which material excitations are made known to us.
My position in regard to the problem of the relation of
the excitation to its sense-manifestation is given in the
introduction to The Mnemic Sensations (p. 4-14). "We
see in an excitation and its manifestation by sensation
not two separate things, but the one thing looked at from
two different points of view." The metaphysical theory
of knowledge does not, however, at this moment concern
us. Throughout, in dealing with the behaviour of organic
substance, I have been concerned exclusively with the
material processes in the phenomena of the higher memory,
of heredity, periodicity, and regulation, and, so fair as I
know, the existence of material changes in the phenomena
of memory is not denied by any scientist.

This is the reason why I adopted a specific terminology
— original excitation, engram, mnemic excitation, etc. —
which has allowed me to be independent of the question
whether the respective material processes manifested
themselves in any given case by conscious sensation or
not. During the whole of the analytical work, as well
as throughout the subsequent synthesis, and especially
in formulating the principal mnemic laws, and in fact
in every definition offered, I avoided the use of such
expressions as memory, remembrance, memory-image, etc

382 THE MNEME

This was done of set purpose, and my reasons for so
doing are set out on page 24. I note, however, on
the part of my critics a determination to involve me in
the use of the word " memory," a term I carefully avoided
both in the analytic and in the synthetic elaboration of
my position. The creation of a special terminology
embodying the main ideas of the theory has justified
itself. Such objections as those of Driesch which really
refer to memory, or to " what is usually called memory
in all systems of psychology," consequently fall to the
ground as inapplicable to the subject actually in question.
Regarding the justification for the simultaneous use
of the subjective and so-called objective methods of
observation, sufficient has been said on pages 40-44. I
think I have there met the objection that Rosenthal
raises when he writes : " One can rightly defend the
position that the logical inferences from data of states
of consciousness can only be applied to data of the same
kind and can never throw light on processes which appear
to us as ' perceptions ' of processes outside of ' Self.' "
If this were indeed true, and if we accepted uncompromis-
ingly the logical consequences of such a position, we should
find that between our own sensations and those of our
fellow creatures there was no reliable link, and that
neither psychological nor physiological research had
any right to establish relations between results obtained
by introspection and those arrived at by the " perception
of processes outside of Self." The practical observance
of the intellectual principles involved in such a position
would be as impossible in daily life as in scientific research,
and as disastrous for experimental as for introspective
psychology — departments of science which, by virtue
of the principles adopted in this book, and notwithstand-
ing the ban of Rosenthal, can look back on splendid
achievements of lasting value.

Let us turn to specific objections against the identity
of the physiological foundation of the phenomena of
reproduction. To Plate, for example, " the main differ-
ence consists in the fact that in memory the recurrence

RETROSPECT 283

takes place in the same brain which received the original
stimulus, while in heredity the recurrence appears in
the next generation ; the problem lies in the transmission
of stimuU from the soma to the germ-cells." But surely
the real difficulty exists only for those who, with Weiss-
mann, deny the possibility of any conduction of excitations
to the germ-cells from a body which is in continuous
connection with them. But, if we may judge by the
statements on pages 328-330 and 344-355 of his book,
Plate does not belong to this group of scientists. The
objection, therefore, sounds strange as coming from him.
Whoever admits somatic induction, that is, a conduction
of the excitations of the soma to the germ-cells, must
accept the possibiUty that traces of these excitations
may be left behind, and consequently the possibility,
where the conducted excitations are sufficiently strong
and the germ-cells in their sensitive or receptive condition,
of corresponding engrams in both germ-ceUs and soma.

For the partisans of an exclusive parallel-induction,
a difficulty exists so far as they have to assume that
stimuli, which directly reach the germ-cells, produce
effects in them without the interposition of the trans-
forming stimulus-receptors of the soma, and that they
leave behind engrams which are of hke nature with those
which are generated by the interposition of complicated
transforming contrivances. From the physiological point
of view, this is inconceivable. For it is by the somatic
reception that the stimulus is transformed into an exci-
tation with distinct local and specific characteristics,
and so enabled to leave behind an engram endowed with
local and specific characteristics. It is in this reception
of the stimulus on the part of the soma that the influences
receive, so to speak, the stamp which, on the induction
of the germ-cells, guarantees in the offspring a recurrence
of locahsed and specific characters.

The question, not very pertinent, is sometimes put :
" How is it that the germ-cell stores up thousands of
engrams, while the ganghon-cell is able to store up only
one, ox at most a very small number of impressions ? "

284 THE MNEME

The reply is simple. The notion that each ganglion-cell
is unable to store up more than a very small number of
impressions implies that the engrams are so locaUsed in
the brain that each cerebral cell represents, so to speak,
a drawer for each single " memory-image." This idea,
held at the beginning of brain-physiological-research, was
soon relinquished as absolutely untenable, and for the
last twenty years or so has been discarded by every
scientific man who specialises in the localisation of cere-
bral function. ■ When, in the first German edition of this
book, I criticised this naive position, friends among brain
physiologists pointed out that the notion was completely
obsolete, and that it would be an anachronism to deal
with it. Accordingly, in the second German edition, I
refrained from referring in detail to such " mythological "
views, as Rieger calls them. If the mere refutation of
such antiquated notions be anachronous, what must be
thought of their apphcation to our attempts to place
the phenomena of organic reproduction on a common
physiological basis ?

Finally, let me say a few words on an objection advanced
by O. Hertwig, who claims that " the material bases of
the brain-substance and of the hereditary matter are
fundamentally distinct," and that, therefore, " the pro-
cesses going on in each are of different kinds." First, I
question whether the , material foundations of brain
substance and hereditary matter are " fundamentally
distinct." For are not these foundations in both cases
nuclear elements — cells? Touching this point, Hertwig
has written : " The distinction of the material foundations
arises from the fact that the single ceU itself manifests
the phenomena of heredity, while the phenomena of
memory are dependent on a nexus between many cells,
that is, on the development of a highly complicated
nervous system and specifically of the cerebral cortex."
But in this Hertwig appears to me to concede that the
difference is not fundamental — that is, a difference in
essence — but that it is simply a difference in complexity.

I adroit that a fundamental difference in, the material

RETROSPECT 285

foundation of these processes would exist if the descriptions
and illustrations of the formation of brain-engrams,
given by H, E. Ziegler, were not hypothetical, but the
results of actual observation. He represents these engrams
as changes in the ramifications of cell-processes — terminals
of the dendrites and neurites — and as formations or
strengthening of neurofibrils. As we have good reason
for the assumption that the engrams of the germ-cells
are largely, if not altogether, locahsed in the cell-nucleus,
and as it has been definitely established that the germ-
cells have neither the tree-like terminals nor the fibrillary
structure which characterise the nerve-cells, a fundamental
difference of the material foundations for the engraphic
changes of the germ-cell and of the nerve-cell would seem
thereby to be given. Unfortunately, no human eye has
as yet seen those engraphic changes of the brain-cells
which Ziegler by word and diagram so eloquently depicts.
They do not possess reahty. In my opinion, the strong
probability is that the actual engraphic changes are in
every respect different from those Ziegler surmises. It is
futile, however, to quarrel over histological problems
which hes so entirely beyond the boundaries of actual
observation, and for which even indirect evidence
is lacking. If Ziegler (p. 38) postulates that every
" physiological explanation must be based on anatomicEil
and histological states," then I reply that it is quite
justifiable, and, in fact, necessary, to treat all faculties
and functions of the organism in their correlations with
corresponding morphological structures, so long as this
is done without straining and without the use of purely
arbitrary hypothesis. But I maintain that enquiry into
this correlation is impracticable as long as an anatomical
basis, ascertained by specific observation, is still wanting ;
for, in the absence of such a basis, one is deaUng only
with statements which, hke the representation of the
brain-engrams by Ziegler, are merely the outcome of
imagination.

At this place I will make a passing reference to
one objection, which is directed against a somewhat

a86 THE MNEME

unimportant point in my evidence, but which has been
advanced by many writers, notably by Ziegler, who
writes : " Semon bases his theory on some of his experi-
ments on plants. The accuracy of his observations,
however, is flatly denied by the famous botanist. Prof.
PfefEer." The readers of this book will be able to judge
how far it is true to say that the Mnemic theory is based
entirely on my experiments on the sleeping movements of
plants. For the evidence derived therefrom is but a frag-
ment of an accumulated mass based on other phenomena
observed and recorded by many workers, such as Chauvin,
Kammerer, Standfuss, Fischer, Schroder, Przibram,
Sumner, Blaringhem, Klebs, Bordage, etc., and which I
have always recognized as far more convincing than my
investigations on the diurnal period of plants, and such
evidence as is not entirely open to experimental verifi-
cation. But, after all, the essential point is that Pfeffer
has in no way denied my conclusions on the only matter
which is here relevant, but on the contrary has confirmed
them, as the reader may see for himself by reference to
the treatise by Pfeffer (see Index).

In conclusion, I may refer to the argument that the
Mneme theory stands in contradiction to MendeUan laws.
If that were true, it would certainly be so much the
worse for my theory. That it is not true has already
been sufficiently shown in Chapter XIV of this work,
wherein it is made clear that our conception of the pheno-
mena of organic reproduction as resting on a common
physiological basis furnishes a necessary complement to
the research work on variation and hybridisation — work
which is concerned with problems other than those dealt
with by the Mnemic theory.

If Mendelian research had actually proved that the
gametes bear either one or the other, but not both, of the
allelomorphic pair of unit-characters, and that the germ-
cell at some point of cell-division in the formation of
gametes divides into two definitely dissimilar parts, and
that the process must be interpreted — as the Mendelians
say— as a " segregration " of characters, then we should

RETROSPECT 387

have to accept this as a fact and draw the necessary
consequence from it, namely, that the separate determining
factors, the engrams, were isolable structures. To me,
however, the avzulable evidence is unconvincing. The
eUmination at the formation of the gametes of one of
the two allelomorphic characters can be explained other-
wise. But I should like to say that the explanation may
be regarded as quite independent of the ideas developed
in this book.

In any case the Mneme theory is not adversely affected
by the Mendehan theory. The incompatibihty of the
two would follow, not from the mere statement of the
possibility of isolation, but from the proving of an actual
isolation existing from the beginning, and from the
demonstration of the lack of orderly connection between
the determinants, that is, the engrams.

By certain modern scientific workers this isolation is
assumed. They see in the hereditary factors a mass of
blindly-shuffled particles, each of which possesses its
special chemical constitution. Because specific factors,
under certain circumstances, cause the formation of
enzymes, they are by some described as enzymes. But,
as I have shown on pages 263-265, that view stands in
contradiction to evidence gathered from the physiology of
development, and from the phenomena of regulation and
of periodicity, and therefore cannot be used as an argument
against the position outlined in this book.

In my judgment, none of the criticisms hitherto raised
has stood the test of examination. I admit, however,
that much work has yet to be done in the experimental
investigation of the principles that govern the somatic
induction of the germ-cell. I do not forget that we are
as yet only at the opening of a new and, as I trust, promis-
ing avenue, and that the main journey hes stiU before
us. If I have succeeded in demonstrating that no obstacle
exists to prevent us from using these fundamental mnemic
laws as a uniform physiological foundation in the investi-
gation of the phenomena of organic reproduction, my
purpose will have been achieved.

CHAPTER XVII

THE MNEME AS THE PERSISTING ELEMENT
IN THE CHANGE OF ORGANIC PHENOMENA

I HAVE now, I hope, given sufficient proof that the
same fundamental physiological laws are valid for all
organic reproduction, and that an identical principle
underlies all organic phenomena which confront us in
such manifold diversity of garb. I now conclude by
asking what we have gained thereby, and how far we
have arrived at a better understanding of the evolved
and ceaselessly evolving organic world in which we live.

The influences in the external world effect changes in
the organism in a twofold manner : first, in the sense
of a merely S5mchronous alteration, secondly, through
and beyond this, by affecting the organism engraphic-
ally and so permanently transforming it. The external
energetic condition, for ever changing, never repeating
itself exactly, thus acts as a transforming factor ; whilst,
in the faculty of the organic substance to retain traces
of the synchronous excitations in the form of engrams,
we have the factor whereby transformations effected
during the " flight of phenomena " are conserved.

But are these two principles sufficient to render the
aspect of the organic world scientifically intelligible to
us? We venture to think they are not. We find the
organisms in a pecuUar, harmonious relation to the outer
world, a relation which has been fitly described as one
of adaptation or adaptabihty to the conditions of life.
Neither the transforming action of the external world
nor the purely conserving mnemic faculty of the organic

a88

THE MNEME THE PERSISTING ELEMENT 289

substance can account for this adaptation. An additional
principle is required to explain it.

The necessity for this has so far been recognised that
two attempts have been made to explain the nature of
this additional principle. First, a mere paraphrase of
the necessary principle was offered in terms of indefinite
or indefinable factors such as " internal causes," or
" impulses " or " desire " on the part of the organism
to adapt itself to such external condition and to develop
in this or that direction. This interpretation seems to
us not only ineffective but also prohibitive of further
enquiry. Lamarck and Nageli may be noted as the most
prominent representatives of this forlorn and barren
explanation.

The second and more adequate attempt to solve this
problem was successfully made by Darwin and Wallace,
who, in my opinion, achieved thereby one of the greatest
triumphs of the human mind. The riddle was solved
by adducing evidence of an operative selective principle
whereby the unfit were eliminated and the fit survived
through succeeding generations.

This foundation laid by Darwin has turned out to be
a thoroughly sohd one, and though it gets amplified
and deepened by the results of more modern experimental
research, it is not shaken, even where corrections have to
be made.

Darwin was not, however, in a position to distinguish,
as we can to-day, the different kinds of variation since
revealed by recent experimental research. He set the
direction and traced the first steps ; further progress had
to be left to his fellow labourers and successors. That
this subsequent work was taken in hand comparatively
late was not his fault ; that new fundamental data were
brought to light by it was only to be expected. The
greatest advances which have been made are due, first,
to the more precise distinction between the inheritable
and non-inheritable characters of the variations, a point
which had engaged Darwin's attention, but which was
hot exhaustively worked out by him ; and secondly, to

19

290 THE MNEME

the discovery that though selection may serve to isolate
already existing types in the mixture of individuals
of a " population," it is unable actually to create new
characters.

New characters are created exclusively by the organic
reaction against stimuli, which continually spring from
the ever changing external world. Darwin recognised
the existence of this organic reaction against stimuli ;
he never denied the direct influence of the externa) world,
and he always guarded himself against such an over-
estimate of the principle of selection as was made by
Weismann under the cry of " The All Sufficiency of
Natural Selection." It is not Darwin, but Weismann,
who has been refuted by the more recent data acquired
in the course of experimental research into the laws of
heredity.

The action of natural selection is indeed but a negative
one, yet it exercises in the change of organic phenomena
an exceedingly effective, and in a limited sense a creative
influence. The operations of a sculptor who chisels a
figure from a block of marble are, it may be said, entirely
of a negative character. Yet just as we see in the group
of Laocoon a creative achievement, so we may regard
many of the astonishing adaptations that confront us
everywhere in organic nature as the shaping work of
Natural Selection, although the latter has just as httle
created its material (the hereditary variations) as the
sculptor the marble out of which he models his creation.

Natural Selection simply weeds out the unfit, but as
the process is ceaselessly at work, we are not surprised
to find everywhere organisms which are adapted to their
environment. In the wonderful phenomena whereby the
adaptation is achieved, not all the elements which in our
generalising terminology we describe as " useful " can
be exclusively traced to Natural Selection. In truth,
these astonishing processes of regeneration and regulation,
to which many antagonists of the doctrine of Selection
appeal, require for their explanation the principle of
Mnemic Persistence, and are not, as Weismann claimed

THE MNEME THE PERSISTING ELEMENT 291

for regeneration, involved in the action of Natural
Selection.

What we are accustomed to describe as " usefulness "
in the organic world is therefore a product of at least two
factors. Natural Selection and Mnemic Persistence. Both
conceptions are the outcome of biological research, and
are based exclusively on the principle of causahty, which
also suffices for the exploration of the inorganic world.
There is no need for " final causes," or " entelechies," and
the hke. The appeal to a vitalistic principle is rendered
superfluous.

Neither in the Mneme nor in Natural Selection is to be
discovered an all-sufficient principle furnishing a universal
key to the understanding of all organic phenomena.
But in the Mneme there is to be found a conserving
principle which is indispensable for organic development,
in so far as it preserves the transformations which the
external world unceasingly creates. Its conserving in-
fluence is, of course, restricted by that indirect factor of
the external world. Natural Selection, for, in the long
run, only fit transformations survive.

This insight into the activity of the Mneme in onto-
genesis supphes also the key to a fuller understanding of
the biogenetic law, in the formulation of which Haeckel
has to a surprising extent deepened and enlarged the
foundations of comparative morphology. That the an-
cestral path of development has to be trodden by each
descendant in an approximately equal manner (PaHno-
genesis) is the obvious consequence of the action of the
mnemic factor in ontogenesis. That in course of time
this path, especially in its oldest and therefore most
frequently traversed initial stages, becomes shortened
here and there and otherwise changed (Caenogenesis) is
equally obvious, if we consider that during each onto-
genesis new original stimuH act on the organism, and
in favourable cases (that is with stimuh of sufficient
strength or of sufficient frequency of repetition in the
lines of generations) add their engraphic effects to the
old mnemic stock.

292 THE MNEME

Moreover, the study of the Mneme is of the greatest
importance, not only for the problems of genetics, but
quite as much for the correct interpretation of the functions
of the living organism.

Restriction to the study of synchronous stimulation
leads to quite a one-sided conception of physiological
phenomena, and, what is worse, to an entire misconception
of the so-called " formative " stimuli. In the great
majority of these it is merely a case of an ecphoric action
of certain stimuli on inherited engrams. Further, the
disappointment ensuing on the realised impossibihty of
understanding many physiological phenomena solely on
the basis of synchronous stimulation leads many thinkers
back to vitalism. This disappointment and this relapse
may be avoided, if the engraphic effects of stimulation
are duly recognised as such, and if it is realised that the
ecphory of inherited and individually acquired engrams
affects many physiological phenomena. We have to
abstain from trying to investigate the physiology of
organisms as if they were isolated from their previous
experiences and those of their ancestors.

We are still far from being able to describe the pheno-
mena of life on a purely physico-chemical basis. But
progress in that direction will be made if we regard
regeneration and regulation, which have hitherto baffled
all mechanistic analysis, as adjusting processes of co-
operating original and mnemic excitations.

By the extension of this conception, certain profound
enigmas of the organic world may become intelhgible
by reference to the mnemic properties of the irritable
substance, a process the main features of which we have
attempted to trace' in the preceding pages.

At the same time, to reach our goal, research work in
physics and chemistry would have to be undertaken in
order to discover whether, and how far, evidence could
be adduced for something corresponding in the inorganic
world to engraphy and ecphory. So far no results in
that direction have been obtained.

Confining ourselves for the present to the investigation

THE MNEME THE PERSISTING ELEMENT 293

of the organic world, we see, on adequate consideration
of engraphic stimulations, how the area of the physiology
of stimuli widens out in an extroardinary manner. We
have found that the phenomena of organic reproduction,
whether of an hereditary or non-hereditary character, are
subject to the same laws, and that they possess the same
physiological basis. Physiology as the science of actual
life cannot dispense with the investigation of the past.
This claim is met by the study of the Mneme, which in
the organic world Unks the past and present in a living
bond.

INDEX OF LITERATURE

CHAPTER I

p. 22. Fr. Darwin and D. F. M. Pertz, On the artificial Production of
Rhythm in Plants. Annals of Botany, vol. xvii, 1903, page 104,

CHAPTER II

p. 28. Francis Darwin, Lectures on the physiology of movement in

plants, i. Associated stimuli. The New Phytologist vol. v.

No. 9, 1906.
p. 29. C. B. Davenport and W. B. Cannon, On the determination of

the direction and rate of movement of organisms by light.

Journ. of Physiol., vol. xxi, 1897, page 32.
p. 29. F. Oltmanns, tjeber positiveu und negativen HeUotropismus.

Flora Oder Allgemeine Botanische Zeitung, vol. 83, 1897.
p. 33. W. Biedermann, Electrophysiologie, Jena, 1895, page loi.
p. 34. E. Steinach, Die Summation einzeln unwirksamer Reize als

allgemeine Lebenserscheinung. Pflugers Archiv fiir Physio-

logie, vol. 125, Heft 5-7, 1908.
p. 47. C. Lloyd Morgan, Habit and instinct. London, New York, 1896.
p. 48. L. Edinger, Haben die Fische ein Gedachtniss ? Miinchen, i899>

Buchdruckerei der " Allgemeinea Zeitung."

CHAPTER III

p. 57. P. Kammerer, Vererbuug erzwungener Fortpflanzungsanpassun-
gen I u. II Mitt, im Archiv fur Entwicklungsmechanik,
vol. 25, 1907.

p. 61. M. von Chauvin, Ueber die Verwandlungsfahigkeit der mexikan-
ischen Axolotl. Zeitschrift fiir wiss. Zoologie, vol. 41, 1885.

p. 61. A. Pictet, Influence d'alimentation et de I'humidit^ sur la varia-
tion des papillons. M6m. Soc. de Phys. et de I'Hist. nat. de
Genfeve, vol. 35, 1905.

p. 61. Chr. Schroder, Ueber experimentell erzeugte Instinktvariationen.
Verhandl. d. D. zool., Gesellschaft, 1903.

p. 6i. P. Kammerer, Vererbuug erzwungener Fortflanzungsanpassungen
III Mitt. Archiv fur Entwicklungsmechanik, vol. 28, 1909.

p. 61. E. Fischer, ExperimenteUe Untersuchungen uber die Vererbuug
erworbener Eigenschaften. Allgemeine Zeitschrift fiir Ento-
mologie, vol. 6, 1901.

p. 61. M. Standfuss, ExperimenteUe zoologische Studien mit Lepidop-
teren. Neue Denkschrift der ailg. Schweiz. Naturforscher
Gesell, vol. 36, 1898 ; also : Zur Frage der Gestaltung und
Vererbuug auf Grund 28 jahriger Experimente. Vortrag,
Zurich, 1905.

29s

296

THE MNEME

p. 63. E. Bordage, A propos de rh£r£dit£ des caracttees acquis..

Bulletin scientifique, 7 s6rie, T. 44. Paris, 1910.
p. 73. W. Pfeffer, Der Einfluss von mechanischer Hemtnung und von
Belastung auf die Schlafbewegungen. Abhandl. d. math-
phys. Klasse der kgl. sachs. Gesellschaft der Wissenschaften,
vol. 32, Leipzig, igii. See also :
R. Stoppel, Zeitschrift fur Botanik, 2 ter Jahrgang, Jena, 19 10.
R. Stoppel und H. Kniep, Zeitschrift fiir Botanik, Jena, igii.
p. 74. W. Pfefier, Biol. Centralblatt, vol. z8, 1908, page 389-415.
p. 74. P. W. Gamble and F. W. Keeble, Hyppolyte varians. Quarterly

Journal Microscop. Science, N.S., vol. 43, 1900.
p. 74. P. W. Gamble and F. W. Keeble, Philosoph. Transact. Royal

Soc. London (B), vol. 196, 1904.
p. 74. G. Bohn, Bull. Institut g6n. psycholog., 3, 1903 ; 7, 1907.
p. 74. W. Schleip, Der Farbenwechsel bei Disdppus morosus. Zool.,
Jahrbuch Abt. f. allg. Zoologie und Physiologic, vol. 30,
Heft I, 1910.
p. 77. Th. H. Morgan, Studies of the " partial " larvae of Sphaerechinus.
Archiv fiir Entw. Mechanik, vol. 2, 1896.
H. Driesch, Neue Antworten und neue Fragen der Entwicke-
lungsphysiologie. Ergebnisse der Anat. und Entw. Gesch.,
vol. II, 1902.
Th. Boveri, Zellenstudien, Heft 5, Jena, 1905.
p. 78. C. Herbst, Formative Reize in der tierischen Ontogenese, Leipzig,

1901, page 59.
p. 78. H. Spemann, Sitzungsbericht der Phys.-med. Gesellschaft, Wurz-
burg, 1901. See further : Anat. Anzeigen, vol. 23, 1903. Zool.
Anzeigen, vol. 28, 1905.
p. 78. E. Mencl, Archiv fiir Entw. Mech., vol. 16, 1903 ; vol. 25, 1908,
p. 78. H. Spemann, Zool. Anzeigen, vol. 31, 1907 ; Verhandlung der
deutschen Zool. Gesellschaft in Rostock und Liibeck, 1907 ;
Verhandlung der Zool. Ges. in Stuttgart, 1908.
p. 78. H. D. King, Archiv fiir Entw. Mech., vol. 19, 1903.
p. 79. W. H. Lewis, American Journal Anat., vol. 3, 1904, and vol. 7,

1907.
p. 79. G. Ekman, Experimentelle Beitrage zum Linsenbildungsproblem
bei den Anuren. Archiv fiir Entwicklungsmechanik, vol. 39,
Heft 2 and 3, 1914.

CHAPTER IV

p. 93. The Life and Letters of Charles Darwin, vol. 1, page 84.

p. 104. E. Wasmann, Die psychologischen Fahigkeiten der Ameisen,

Stuttgart, 1899.
p. 105. H. Ebbinghaus, Ueber das Gedachtnis. Untersuchungen zur

experimentellen Psychologie, Leipzig, 1885.
p. 108. J. Fabre, Souvenirs entomologiques, Paris, 1879-1882.

CHAPTER V

p. 128. C. von Monakow, Ueber die Lokalisation der Hirnfunctionen.

Wiesbaden, E. F. Bergmann, 1910.
p. 130. Marie von Chauvin, Ueber die Verwandlungsfahigkeit des mexi-

kanischen Axolotl. Zeitschrift fur wissensch. Zoologie, vol. 41,

1885.

INDEX OF LITERATURE 29;-

p. 230. p. Kammerer, Vererbung erzwungener Fortpflanzungsanpas-
sungen. 3. Mitteil. Archiv fiir Entw. Mechanik, vol. 28,
1908. page 526. Also Zeitschrift fur induktive Abstammungs-
und Vererbungslehre, vol. i, 1909, page 139.

p. 132. P. Kammerer, Direkt induzierte Farbanpassungen und deren
Vererbung Zeitschrift fiir induktive Abstammungs und Verer-
bungslehre, vol. 4, Heft 3 and 4, 191 1.

p. 132. H. Frzibram, Aufzucht Farbwechsel und Regeneration der
Gottesanbeterin (Mantidae) III. Archiv fiir Entw. Mech.,
vol. 28, 1909. See also the same author's Experimentelle-
Zoologie, vol. 3, Phylogenese, Leipzig and Wien, 1910, page>
161.

p. 133. C. Detto, Die Theorie der direkten Anpassung, Jena, 1904.

p. 133. A. Weismann, Aufsatze iiber Vererbung, Jena, 1892, page 323.

p. 134. F. B. Sumner, An experimental study of somatic modifications,
and their reappearance in the offspring. Arch. f. Entw.-
Mechanik, vol. 30, II Teil, 1910.

CHAPTER IX

p. 193. J. Loeb, The dynamics of living matter. New York, 1906, page 171.-
P- 193- J- I-oeb, Ueber das Wesen der formativen Reizung, BerUn, 1909.
P- 193- J- Loeb, Die chemische Entwicklungserregung des tierischen
Eies. (Kiinstliche Parthenogenese), Berlin, 1909.

CHAPTER XI

p. 210. Th. Boveri, Ueber die Polaritat des Seeigel-Eies. Verhandl. der
Phys.-Med. Ges., Wurzburg N.F., vol. 34, 1901.

p. 211. H. Driesch, Neue Erganzungen zur Entwickelungsphysiologie
des Echinidenkeimes. . Archiv f. Ent'wicklungsmechan, vol. 14,.
1902.

p. 213. P. Kammerer, Centralblatt fiir Physiologic, vol. 19, 1905.

p. 214. P. Kammerer, Archiv f. Entw. Mech., vol. 19, Heft 2, 1905.

p. 219. H. Przibram, Experimental Zoologie, vol. II, Regeneration,
pages 1 1 7-1 1 9, Leipzig und Berlin, 1909. Also : Die Homoeosis
bei Arthropden. Archiv fur Ent-wicklungsmech., vol. 29, 1910..

CHAPTER XII

p. 223. Klein, Futterbrei und weibliche Bienenlarve. " Die Bienen-
pflege," 26 Jahrg,, Ludwigsburg, 1904, page 80.

p. 223. H. V. Buttel-Reepen, Atavistische Erscheinungen im Bienen-
staat. C.R.d.1., Congres intemat. d'entomologie k Bruxelles,
1910; Briissel, 1911.

p. 225. A Forel, Les fourmis de la Suisse. Neue Denkschrift d. allg..
Schweiz. Ges, f.d. allg. Naturw., vol. 26, 1874.

p. 225. E. Wasmann, Ergatogyne Formen bei den Ameisen. Biologisches-
Centralblatt, vol. 15, 1895.

p. 225. H. Viehmeyer, Experimente zu Wasmanns Pseudogynen-
Lomechusa Theorie Allgem. Zeitschr. f. Entom., vol. 9, 1904..

p. 226. H. V. Buttel-Reepen, Die stammesgeschichtliche Entstehung
des Bienenstaates. Biol. Centralblatt, vol. 23, 1903. Pub-
lished also separately with additions by G. Thieme, Leipzig,
1903.

p. 230. Marie v. Chauvin, Zeitschrift f. wissenschaftl. Zoologie, 1875,.
1876, 1885.

:298 THE MNEME

p. 236^ p. Kammerer, Vererbung erzwungener Fortpflanzungsanpaa-
sungen III Mitt. Die Nachkommen der nicht brutpflegenden
Alytes obstetricans. Archiv. f. Entwicklungsmechanik, vol. 28,
1909.

CHAPTER XIII

■p. 241. H. Przibram, Experimentelle Studien fiber Regeneration. Archiv
fur Entwicklungsmech, vol. 11, 1901.. Also Experimental-
Zoologie. 2. Regeneration, Leipzig and Wien, 1909, p. 108.

p. 242. H. Nilsson-Ehle, Ueber Falle spontanen Wegfallens eines Hem-,
mungsfactors beim Hafer. Zeitschr. f. induktive Abstammungs-
und Vererbungslehre, vol. 5, Heft i, April, 191 1.

p. 244. E. v. Tschermak, Ueber Zuchtung neuer Getreiderassen mittels
kiinstlicher Kreuzung, II Mitt. Zeitschr. f.d. landw. Versuch-
swesen in Oesterreich, 1906.

p. 244. Vernon, Archiv f. Entw. Mech., vol. 9, 1900.

■p. 244. Doncaster, Philosoph. transact., vol. 196, 1903.

p. 244. Herbst, Vererbungsstudien, IV, V, VI. Archiv f. Entw. Mech.,
vol. 22, 1906 ; vol. 24, 1907 ; vol. 27, 1910.

p. 244. Tennent, Archiv f. Entw. Mech., vol. 29, 1910.

p. 244. Tower, Biological Bulletin, vol. 18, No. 6, 1910.

"p. 253. W. L. Tower, The determination of dominance and the modi-
ficatiou of behaviour in alternative (Mendelian) inheritance
by conditions surrounding or incident upon the germ-cells
at fertilisation. Biological bulletin, vol. 18, No. 6, 1910.

CHAPTER XIV

,p. 257, W. L. Tower, An investigation of evolution in chrysomelid

beetles of the genus Leptinotarsa, Carnegie Institution,

Washington, 1906.
p. 257. P. Kammerer, Mendelsche Regeln und Erwerbung erworbener

Eigenschaften Verhandl. d. naturf, Vereines in Brunn, vol.

49, 191 1. Further: P. Kammerer, Vererbung erworbener

Farbenveranderungen. IV Mitt. Das Farbkleid des Feuer-

salamanders in seiner Abhangigkeit von der Umwelt. Archiv

f. Entwicklungsmechanik, vol. 36, 1913.
p. 257. Charles Darwin, The foundations of the origin of species. Two

essays written in 1842 and 1844. Edited by his son, Francis

Darwin,
p. 25S. H. de Vries, Die Mutationstheorie, Leipzig, 1901-1903.
p. 258. W. L. Tower, The determination of dominance. Biological

Bulletin, vol. 18, No. 6, 1910.
p. 259. F. B. Sumner, An experimental study of somatic modifications

and their reappearance in the offspring. Archiv f. Entw.-

Mech., vol. 30, II Teil, 1910.
p. 259. H. Przibram, Versuche an Hitzeratten. Verhdlg d. Ges. deutscher

Naturforscher und Aerzte. Versammlung, Salzburg, 1909.
-p. 261. H. Nilsson-Ehle, Kreuzungsuntersuchungen an Hafer und

Weizen. Lands Universitets Aersskrift N.F. Afd. 2, vol. 5,

No. 2, Lund, 1909.
T). 263. W. Bateson, Mendel's Principles of Heredity. Cambridge, 1909,

chap. ix.
p. 263. H. Nilsson-Ehle, Ueber FaUe spontanen Wegfallens eines Hem-

mungsfaktors beim Hafer. Zeitschr. f. induktive Abstam-

mungs-und Vererbungslehre, vol. 5, Heft i, 191 1.
p. 265. W. L. Tower, The determination of dominance. Biological

Bulletin, vol. xviii. No. 6, 1910, page 324.

INDEX OF LITERATURE 299

CHAPTER XVI

P- 275. J. R. Mayer, Bemerkungen fiber das Aequivalent der Warme,
1850. " When a phenomenon has become known in all its
aspects, it is already explained thereby and the task of
science is complete."

p. 280. O. Hertwig, Die Zelle und die Gewebe, vol. 2, Jena, 1898, pp.
245 and 251, new editions under the title : Allgemeine Biologie,
2nd edition, Jena, 1906, p. 587; 3rd edition, Jena, 1909, p. 661.

p. 280. L. Plate, Archiv f. Rassen und Gesellschaftsbiologie, 6 Jahrg.
Leipzig and Berlin, 1909, page 92. {Review of Francis Darwin's
Presidential Address, Dublin, igo8.)

p. 281. L. Plate, Selektionsprincip und Probleme der Artbildung, 3
Aufiage, Leipzig, 1908, page 334.

p. 282. H. Driesch, Philosophie des Organischen, Leipzig, 1909, vol. i,
pages 220-223.

p. 282. J. Rosenthal, Biologisclies Centralblatt, vol. 25, 1905, page 368,
in a critical review of the Mneme.

p. 282. L. Plate, Selektionsprincip und iProbleme der Artbildung, 3
Auflage, Leipzig, 1908, page 335.

p. 284. O. Hertwig, Die Zelle und die Gewebe, vol. 2, Jena, 1898, page 251.

p. 284. O. Hertwig, Allgemeine Biologie, 3 Auflage, 1909, p. 661.

p. 285. H. E. Ziegler, Theoretisches zur Tierpsychologie und verglei-
chender Neurophysiologie. Biologisches Centralblatt, vol. 20,
1900. The hypothetical descriptions and illustrations also
reappear in the second edition of Ziegler's essay : Der Begriff
des Instincts einst und jetzt. Jena, 1910, p. 88, Figs. 8 and 9.

Richard Semon, Die Mneme, Leipzig, Wilhelm Engelmann,
Erste Auflage, 1904 ; Zweite Auflage, 1908 ; Dritte Auflage,
1911.

INDEX OF AUTHORS

Aikins, H. A., 28

Barfurth, D., 213, 215
Biedermann, W., 33, 34
Blaringhem, L., 85, 175, 256, 286
Bohn, G., 28, 74
Bordage, E., 63, 64, 85, 175, 180,

256, 286
Butler, S., 10

Buttel-Reepen, H. v., 223, 224
Boveri, Th., 77, 210, 211

Cannon, W. B., 29, 44

Charbonnier, 70

Chauvin, M. v., 61, 85, 130, 131,

146, 175, 180, 230, 232, 234,

235. 256, 286
Claypole, 70
Conrad, 253
Crampton, H. E., 211

Darwin, Ch., 93, 258, 289, 290

Darwin, Fr., 22, 28

Davenport, Ch. B,, 29, 44

Detto, C, 133

Doncaster, L., 244

Driesch, H., 77, 190, 211, 268, 282

Dum^ril, A. M. C, 230

Ebbinghaus, H., 105, 106, 107
Edinger, L., 48
Ekman, G. E., 79
Escherich, K., 227

Fabre, J., 109

Fielde, A. M., 227

Fischer, E., 61, 62, 85, 175, 180,

256, 286
Forel, A., 104, 225, 226

Gamble, P. W., 74
Ghinst, van der, 28
Goldstein, K., 215, 216
Goltz, F., 127
Grassi, B., 227

Haeckel, E., 291
Harrison, R. G., 215, 216
Herbst, C., 78, 218, 219, 244
Heiing, E., 9, 41
Hertwig, O., 280, 284
Hodge, C. F., 28
Huber, P., 71

Janet, Ch., 225
Jennings, H. S., 28

Kammerer, P., 58, 60, 61, 85, 132',.
175, 180, 213, 236, 238, 256, 286-
Keeble, F. W., 74
King, H. D., 78
Klebs, G., 85, 175, 256, 286
Klein, 223, 224

Lamarck, J., 289

Laycock, F., 10

Lewis, W. H., 79

Loeb, J., 192, 193, 194, 215

Lubbock, J., 227

Macnish, 144
Mencl, E., 78
Mendel, Gr., 85, 243, 245-248, 251,,

255
Metschnikofi, E., 268
Monakow, C. v., 128
Morgan, C. Lloyd, 47, 70
Morgan, Th. H., 77, 268
MuUer, J., 41

Nageli, C, 289

Nilsson-Ehle, H., 242, 261, 262, 263;

Oltmanns, F., 29, 44
Orr, H. D., 10

Pertz, D. F. M., 22, 28
Pfefifer, W., 38, 73, 286
Philippeaux, T. M., 218
Pictet, A., 61, 63, 85, 175, 256
Planta, 224

300

INDEX OF AUTHORS

301

Plate, L., 280-283
Przibram, H., 85, 132, 134, 175, Z19.
241, 256, 259, 286

Raffaele, 215
Reichenbach, H., 227
Ribot, Th., 144
Rieger, K., 284
Rosenthal, J., 282
Roux, W., 147
Rubin, R., 215

"Schaper, A., 215, 216

Schleip, W., 74

-Schrader, M. E. G., 127

Schroder, Chr., 61, 85, 175, 256, 286

Schultze, O., 195
Selenka, E., 75

Silvestri, F., 227
Spallanzani, L., 213
'Spemann, H., 78, 79

Spencer, Herbert, 226
^Standfuss, M., 61, 85, 175, 180, 256,
286

Steinach, E., 34

Sumner, F. B., 85, 134, 175, 256,
259, 286

Tanner, 227
Tennent, D. H., 244
Tower, W. L., 61, 85, 118, 135, 175,
180, 244, 252-254, 256-258, 265
Tschermak, E. v., 244

Vernon, H. N., 244
,Viehmeyer, H., 225, 227
Vries, H. de., 254, 258

Wallace, A. R., 289
Wasmann, E., 104, 225, 227
Weismann, A., 133, 230, 283, 290
Wheeler, W. M., 225, 227
Wohlgemuth, A., 107

Ziegler, H. E., 285, 286

GENERAL INDEX

Abramis, 49

Abstraction of homophony, 164

Acacia, 73, 147

Acraina, 268

Actinia, 74

Albinotic aberration, 63

Albizzia, 73

Alytes, 61, 236

Amblystoma, 130, 146, 230-235

Ambulatory limbs, 241

Amnesia, periodic, 144

Amphibia, 177, 208

Amphioxus, 177, 208

Andalusian fowl, 243, 246

Annelida, 176, 2IO

Ants, 49, 105, 225

Anura, 213, 214

Aphasia, amnesic, 119

Aphides, igg

Apis mellifica, 222, 227

Arctia caja, 61

Area proper of excitation, 121,

122, 220
Ascidia, 176
Atemeles, 225
Atrophy, 216
Auricula, 51
Axolotl, 130, 146, 196, 230-235

Beech, 51, 52, 55, 147
Bees, 49
Begonia, 114
Biogenetic law, 291
Bombinator, 78, 79
Breeding pads, 238

Caenogenesis, 291
Capri, 36, 92, 143, 174
Cataract, 252
Caterpillar, 71, 81
Cephalopodes, 137
Cerebral cortex, 119, 126
Cerebral ganglion, 137
Chiasma, 126
Chronogeneous engram, 55

Chronogeneous ecphory, 55, 69,

72, 80
Chronogeneous localisation, 276
Colorado beetles, 244, 253, 257
Colour-blindness, 252
Cones, 91

Corpora quadrigemina, 126
Crayfish, 241
Crocus, 51, 52
Crustacea, 28, 74, 195, 20S, 218,.

219
Ctenophores, 176, 211
Cuttlefishes, 49
Cyclic changes, 203, 205
Cyt:oplasmic localisation, 211

Daphne, 51

Daphniae, 29, 44

Decrease of regenerative capacity,

209
Dentalium, 210
Dionsea, 33

Diurnal periodicity, 73, 205, 286
Dixippus, 74
Doe, 156

Dog, 26, 47, 127, 156
Dominance, 243, 244
Double personality, 144

Ebb and tide, 74

Echeneis, 48

Echinidae, 189, 211

Echinoderms, 66, 176, 177, 208,.

268
Entelechy, 277, 291
Euchelia, 48
Euplocamus, 253
Eye (regeneration), 218

Fagus silvatica, 52

Falcon, 127

Ferns, 195

Fishes, 48

Fissure parieto-occipitalis.^t 19

Flagellates, 34, 84

302

GENERAL INDEX

303.

Forcing of plants, 51, 72
Formica, 225
Fox terrier, 156

Galathea, 208
Galloway cattle, 228
Ganglion opticum, 218
Garneele, 241
Gastrulation, 75, 77, 190
Geniculate body, 126
Gracilaria, 61
Guinea fowl, 47

Hair growth, 135

Hares, 253

Heliotaxis, 161

Heliotropism, 84, i6i

HescU's convolution, 129

Heterochely, 220

Heterochrony, 77

Heteromorphosis, 219

Heterozygotes, 250

Hippolyte, 74

Homozygotes, 250

Honey bee, 222, 224, 226,

227
Humble bee, 226
Hybridisation-research, 262, 264,

286
Hydra, 144, 207, 212
Hydromedusae, 177
Hydrophobia, 121
Hymenoptera, 137

Idus, 49

Ilyanassa, 210

Induction, parallel and somatic,

133, 256, 278, 279, 283
Infusoria, 34, 114
Insects, 195
Intoxication, 144
Isolation of conduits, 121, 122

Lanice, 210

Lasius, 105

Lathyrus, 243

Lens of eye, 77, 78, 173

Leontodon, 114

LeporidsE, 253

Leptinotarsa, 118, 244, 253, 257

Leucojum, 51, 52

Ligule, 261

Lilacs, 52

Liverworts, 38

Localisation of symptoms, 128

Lomechusa, 225

Lonicera, 51, 52

Lymantria, 6i

Magpie, 70
Mantis, 133
Matthiola, 243
Maxilliped, 241
Medusae, 208, 268
Melanotic aberration, 63
Membrane formation, 192-194
Memory-image, 24, 119, 280
Mice, 259

Migration impulse, 72
Mimosa, 30, 35, 147
Mirabilis jalapa, 243
Mnestic processes, 129
MoUuscae, 74, 176, 210
Myrmica, 225
Myzostoma, 210

Neoplastic capacity, 215
Nereis, 210
Nest-building, 157

Oats, 261

Obstetric toad, 236, 241
Occipital lobes, 126
Ocneria, 61
Oenothera, 254, 258
Opening tetanus, 22
Ostrich, 70

Paedogenesis, 29

Palcsmon squilla, 74

Palingenesis, 291

Palolo worm, 205

Parthenogenesis, 192

Patella, 210

Peach tree, 64

Pharyngeal ganglion, 137

Pharynx, 269

Phasogeneous ecphory, 56, 69, 79,,

80
Pheasant, 47, 70
Phratora, 61
Pigeons, 127
Pisum, 245
Planaria, 114, 207
Pleiotype, 253
Porcellana, 208
Portunus, 208
Praying mantis, 133
Premature parturition, 61, 174
Protomer, 116, 124, 136
Pseudogynes, 225

Rabbits, 253
Radium radiation, 32
Rana esculenta, 78
Rana fusca, 78, 195
Rana palustris, -j^j^^

304

Sana silvatica, 78, 79
Rats, 259

Receptor of stimuli, 133
Reflex spasms, 123
Reunion, 64
Rods, 91
Rotatoria, 195
Rust, 252

Salamandra, 57, 58, 60, 132, 174,

201, 235
Salamandrina, 198, 213, 214, 235
Scilla, 51
•Scorzonera, 114
Sea-urchins, 244
Sebaceous glands, 134, 135
Sensible period of germ-cells, 118,

135. 257
Short fingers, 252
Silver pheasant, 253
Siredon, 130, 214, 216, 230-235
Sldn, 259
Snails, 74

"Southdown sheep, 229
Spectre locust, 74, 92
Sphere of hearing, 129
Sphex, 108, 109
Spurious alielomorphism, 263
Stag, 156
Starfishes, 195
Stentor, 114, 115
Sterility, 226
.Strongylocentroius, 210, 211

THE MNEME

Strychnine poisoning, 121
Subliminal value, 33
Suffolk cattle, 228
Synapta, 75

Tadpole, 213, 236
Teleostians, 47, 177, 208
Temporal convolutions, 129
Temporal lobes, 119
Termites, 226, 227
Tetanus, 121
Thalamus, 126
Thigmomorphosis, 173
Topochemical sense of smell, 104
Trichotomy, no
Triton, 196, 214, 235
Trout embryo, 213

Urodela?, 213, 214, 218

Vagus fibres, 122
Vicarious ecphory, 45, 70
Vorticella, 28

Wasps, 49
Waterfowl, 47
Weaver birds, 209
Wheat, 261
Worms, 74

Xenodus, 225

Zytolosis, 193

Printed in Great Britain iy
47NWIN BROTHERS, LIMITED, THE GRESIIAM PKESS, WOKING USSD LONDOX